https://www.chadco.co/home daily 1.0 2024-05-21 https://www.chadco.co/bio daily 0.75 2022-04-18 https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594666771045-VO7JX3KJSNK0EODSFGQB/masked+man+profile+pic.PNG Bio - so, we meet again! Below, there’s a whole lot more about me and how I approach my work. Feel free to browse, or just click the shiny button below to say ‘hi’ and talk to me directly. https://www.chadco.co/wx-voice daily 0.75 2020-07-22 https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594255769055-OTLBLV3JBSOAGEHHEBQQ/What%27s+the+WX.png 01. The Weather Channel on Voice Right now, in Atlanta, Georgia, it’s 68 degrees and mostly cloudy. Later, expect a high temperature of 80 degrees and a low of 63. Chance of rain is 0%. https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1595174846348-NT0LWZNVTN1RW8J3NPL2/carlos-macias-JrPX9GD1VPk-unsplash.jpg 01. The Weather Channel on Voice - User-centered design. In parallel to market and technical analysis, I set out on a path to understand our user. The process resulted in three core personas. But, since smart speakers were still a relatively new technology, I considered some elements to guide not only 1:1 interviews but design thinking activities to really help me and the team understand how these devices are used generally and where user needs around weather might not be fully met by existing apps. https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1595175287798-243G3UA3VG0M2VJT5JB0/Screen%2BShot%2B2020-07-19%2Bat%2B11.12.22%2BAM.jpg 01. The Weather Channel on Voice https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1595175797765-N1ZYNMHN3H1B8FB92UDV/floor+plan.001.png 01. The Weather Channel on Voice https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1595176961538-1ISD743ZD3U37V0K8TR3/Screen+Shot+2020-07-19+at+11.42.20+AM.png 01. The Weather Channel on Voice https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1595177312769-369AVALKGUMSPG08SWRF/Screen%252BShot%252B2020-07-19%252Bat%252B11.45.01%252BAM.jpg 01. The Weather Channel on Voice - Id intent. Without a data scientist or linguist on the team, I orchestrated a thorough review of the information architecture of TWC’s digital properties identify topics of conversation. In this case, there were general topics like ‘weather’ and ‘precipitation’ and more specific topics like ‘weather at a specific time today.’ Looking at these general categories and their variations gave us intents and sub-intents so that we could begin exploratory development and begin creating dialogue flows. Later, we would approach the process of identifying questions differently. Partnering with research, we constructed a simple survey using a Jeopardy model. We gave users an answer and asked them to tell us what question they think they’d ask. So, I was able to review and code results to understand what words a user would use to to train the assistants to handle users questions. https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594255283451-3DQ108MVVD70SZPPN597/MVP%2Bscript.jpg 01. The Weather Channel on Voice - Gone in 20 seconds. Based on of user research and design thinking outputs, I wrote sample dialogues that covered a variety of weather conditions, including sunny day, rainy day, severe storms, etc. I got the team together for table reads. Each time, asking a different team member to read the ‘user’ lines and another team member to read Alexa’s lines. I also developed a ‘cue card’ system. For the cue cards, I wanted a system where ‘the user’ couldn’t read ahead and know what the bot was going to say. So, the team member playing the user would look at a cue card and read the user’s line. Another team member would respond as the bot’. Afterward, we discussed what we heard, what was easy to understand, what was hard to understand, what data points stood out, and which ones might be missing. We also extended this system of table reads to the IBM network as a kind of ‘guerilla’ prototype test. I recruited from across the US IBM network. On a video call, I asked some warm-up questions to understand their habits on weather, smart speakers, and other voice assistants. Then, I pulled out the the cue cards. . “It takes too long,” was the common response. “It takes longer than it would take me to look at my phone." Participants liked the information and thought it was useful but didn’t’t want to have to have a conversation to get it. They said they liked the note about humidity and that other apps or skills they use didn’t surface any data that might help them understand how the weather might feel Since a conversational interaction would put off users, I needed a way to satisfy the user need through a command-and-response model. My next round of designs had to condense the information in this dialogue to a single response, hit all the highlights, be read / heard in about 20 seconds, and contain extras like the humidity insight. https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594255043852-MR3T36PBTETILDKQEK0D/WX+custom.png 01. The Weather Channel on Voice - The DIY approach In a way, we had a leg up. Through our relationship with the cable network, our weather sciences and technology teams had developed solutions years ago to make videos that displayed on TV. These updates were created to provide text for 24-hour and 12-hour dayparts that a TTS engine could process and delivery on top of a video. Looking at these, I had some doubts about whether these were a fit for this use case, given what information was or wasn’t included. So, I made a new one to have a third option. To construct a new response, I looked at the dialogue flow and removed all text surrounding the data points. This left only the data points that come from an API. With only a current temperature, a high temperature, a low temperature, a weather condition (i.e. partly cloudy, rain, sunny, etc.), and a chance of rain, I wrote sample responses for a sunny day and a rainy day. In reviewing the developers and other team members, I learned that a separate API can deliver precipitation accumulation. We added this to the rainy day answer, since it seemed important. Lastly, I knew we needed something extra to provide value to the user. So, I identified five dominant conditions (humidity, wind, UV, pollen, visibility, that, if too high or low, could alter how the user might feel or plan for their day. These, I asked the developers to add to our messages as insights. I also led the analysis and helped define business rules to rank conditions, since we were likely only going to be able to return one insight per response. https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1595178394322-K0L6DLTZDOAYTWRR1ZNJ/Screen+Shot+2020-07-06+at+6.59.32+PM.png 01. The Weather Channel on Voice - Being Extra. To get to ‘something extra,’ I deepened my analysis of available data. I looked at all of the API documentation. Through this process, I identified a class of serviceable insights that could be added to our responses. I identified seven ‘dominant conditions,’ that, if present in signficant amounts, would effect how the user might feel or navigate the world on a given day. And, I identified other insight categories like signficant temperature change or ‘flux’ messages that might also important. Then, working with the meteorologists, I analyst the data values for things like UV index to understand what values make it significant and how to communicate that to users and give it context and meaning in terms of a weather report. https://www.chadco.co/severe daily 0.75 2022-04-18 https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594683511906-91WQI9KJBDDG2190PGFB/severe+script.png 03. Severe Weather - An alert from The National Weather Service: A National Weather Service (NWS) alert can be anything from a Small Craft advisory to a hurricane or tornado warning. Communicating these to users is then totally straight-forward, because you don’t want to give something that a general user might not understand (i.e. a small craft advisory) the same weight as a tornado warning. These alerts can be life-saving. I worked with the developer to put the most severe alerts (i.e. tornado or hurricane warnings) first. So that the user hears the most important piece of information first, before the app moves on to answering the user’s question. Less urgent alerts, like a small craft alert, are played after the user’s question is answered. In all cases, users are given the chance to hear the full alert message. From a quality perspective, this isn’t so ideal, because The Weather Channel, like all weather providers, gets this data from the NWS with no ability to format or rewrite it. If the user says ‘yes’ to hear the alert, it will cause the bot to read the full statement from the NWS. Sometimes, these are well-written and easy for a voice assistant to read. At other times, they’ve got a sloppy sentence structure and / or weird punctation. This will trip up Alexa and basically any other voice assistant. But, users in testing told us that that’s ok. “I can always tell Alexa to stop,” said one user. https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594684351274-L4XCS1U9Y844QQMF7WYE/pre.jpg 03. Severe Weather https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594684292356-0EB3UPW0OYM5JS5FIGG8/big%2Bideas.jpg 03. Severe Weather https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594684400061-QL83XHO5X3HUE46Z6CM3/prioritization.jpg 03. Severe Weather https://www.chadco.co/hele daily 0.75 2022-04-18 https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1595435233586-9N9ZT59KC5QGLY7JENQP/phone_mockup_2%403x.png 06. Hele - Conversation-first design Named from the Hawaiian word for trip, Hele is the solution for the modern explorer. A traveler who has an open mind, is curious, and wants to see and experience as much of the world as possible. Hele is a conversation driven interface designed to make travel planning simple and fun. While other travel tools have improved, planning a trip can still be a frustrating experience, requiring multiple tools and time-consuming research for users to make an informed decision. We know that leisure travelers plan trips as far out as 6 months in advance. But do we know and build tools around their most frequent considerations in planning a trip? The top considerations are: “Bucket List” (destination), price, and weather. In fact, 47% of users consider weather before making a purchase decision for travel. So, Hele gives the user a streamlined tool to plan and purchase travel. To get started, a user only has to tell where or when they want to travel. Or, chose to be surprised by clicking “surprise me.” We do the rest of the work for users by combining semi-public flight data, weather, and artificial intelligence (AI) to balance their concerns on destination, budget, and weather. Giving users an opportunity to find the best chance of experiencing a new or favorite place under the best possible conditions. https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1595434548519-YFWL0GT1USNN1EUGA57C/hele%2Bdesign.jpg 06. Hele - The question at its core The 2019 Weather Channel hackathon asked participants to deliver new product concepts built on weather data. In an early whiteboard session, I led the team in identifying criteria that a user might consider when planning a trip. Considering these criteria and how much weather plays into travel planning, the team settled on a destination finder. I then proposed flows based upon two user entry points: I know where I want to go, but when is the best time to go? I know when I want to go, but where I can go and have good weather? https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1595434813555-8X5P1AU7BUNURPO4NAEN/Screen%252BShot%252B2020-07-13%252Bat%252B5.55.23%252BPM.jpg 06. Hele - Sponsorship Potential Certainly, the entire app could certainly be licensed (and then customized to fit any brand. Given all of the open space in the visual design, sponsorship opportunities in the form of contextually relevant ads are ready-made. It’s easy to see where a partner could takeover one or all of the screens throughout the flow. Through the ‘Surprise Me’ flow, an opportunity for partners like Disney Resorts, Carnival Cruises, etc. to show users package or other deals to entice sales. But, in true to its conversation-led structure, it’s easy to image either sponsored intents or sponsored utterances. So instead of just a visual takeover, a partner like Southwest could enter the conversation. It could be branded utterances or responses, or even a branded voice that gets a brand into the conversation and under consideration while users are browsing and making travel decisions. https://www.chadco.co/pew daily 0.75 2020-07-13 https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594662044005-88QWRGX7TCMCTB4KLC4N/Screen+Shot+2020-07-13+at+12.39.58+PM.png 05. The Flip Side - Extending the reach of great content When the Pew Charitable Trust approached us, they were looking for new ways to put their content in front of a bigger audience. While their insights weren’t a fit for The Weather Channel’s voice apps, they are highly informative, surprising, and create wow or aha! moments. I remembered that , during testing, users had told me they look to voice assistants and conversational interfaces to learn something new. What better way to help Pew share data and give users access to new information than to modify Pew’s existing concept and put it out as a Flash Briefing for Alexa or a news feature for the Google assistant? https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594662801203-1GLDPDJAS8I9PS8L9JIH/social+media+and+news.png 05. The Flip Side - The Social News. The content was great and it was well written. But, it need a little something to alert the user that there’s something interesting coming. So, I reworked and edited existing content to give it a hook. Here, the existing article mostly worked. So, I wrote the introduction and left the original article mostly untouched. I think the body text could be edited for brevity and SSML tags might improve comprehension. But, all in all, I think this example represents the potential of Pew’s content on voice. https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594664046944-JMD3ZX5QYYMMYNWERC04/Pew+2.png 05. The Flip Side - Gen X & the Great Recession While the first concept was a great fit, others were more challenging. To show Pew, I adapted an article about Generation X , wealth, and the great recession to this news briefing concept. The hook worked beautifully, but the information is hard to digest, since there are so many numbers used as points of comparison. I did copy edit a bit for them, but I didn’t do a full rewrite. My point was to show them how the might think of ‘dryer’ more research-oriented content within this format. I also provided them with guidance on how to think about copy in the context of audio and introduced concepts like table reads to help them think about how users might hear such content and how to write for an audio format. https://www.chadco.co/future-wx daily 0.75 2020-07-15 https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594846015651-S0XOOR79V7Z347YUDJAC/Screen%2BShot%2B2020-07-15%2Bat%2B3.45.10%2BPM.jpg 04. The Future of Weather on Voice - Take it to the app. What is Alexa but a chatbot with a voice interface? I directed our teams to architect to deliver a conversational experience across multiple voice platforms and through chat. These concepts give The Weather Channel (TWC) as a brand the option to audio-ize alerts, incorporate national weather stories into a user’s routines, and offer chat. Thus, deepening a user’s engagement and get the weather how they want to get it https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594658800699-V8IQRQ755916751J9JIG/com4com.png 04. The Future of Weather on Voice The user doesn’t care how it gets done. So, broker a B2B2C experience that solves a user need. https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594658706008-EZ2TUQNS3L769T2SD3T7/vx+takeover.png 04. The Future of Weather on Voice “It’s 64 degrees where you are & YOU GET A CAR!” (j/k ) But… Useing celebrity and other branded voices can make getting the weather fun and provide ‘ad’ revenue. https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594658907288-X898KXUYP88GI5B0HRS8/promo+codes.png 04. The Future of Weather on Voice Simple to executes. Users love a deal. And, it’s trackable in most ecommerce platforms. https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594661334889-EVTFYCPL83MIVNHJ4JXH/OTT+2.png 04. The Future of Weather on Voice - Use all of that screen! Especially when it might be 80 inches (>200 cm)!! And, most OTT systems offer voice control Heck, to compete up with Xfinity, AT&T is even marketing it’s streaming device by showing how easy it is to use voice to find something to watch. But what about publishers and other content providers? https://www.chadco.co/testing daily 0.75 2020-07-16 https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594321866220-W5P1C6GIA3CF2MPQZ25C/UZ.png 02. Content Testing at Scale - Method, not Madness In partnership with UserZoom and TWC research, I was part of an industry first effort to test smart speaker content at scale. UserZoom provided the technology and worked with research on the methodology, ultimately borrowing from the CPG vertical to use a JAR or “just about right” scale. In this methodology, the user is presented with a scale to tell us whether the response has too little, too much, or is just about right. Foe these tests, success is >65% replying with a ‘3’. For my part, I identified all of the required tests and sorted them into epics. I wrote copy variations and recorded the necessary files for use in the surveys. To maintain the required pace, I ran surveys and documented and interpreted results. Ultimately, our response consistently beat our nearest competitors, the default Alexa weather skill and the Big Sky skill. I also drove optimizations on presentation of all data elements and insight types to ensure TWC met user needs and wants at launch. https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594934507294-XP66OYP6EHZKJA39W0A0/twc+12+hr.png 02. Content Testing at Scale https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594934478445-S6CMQI2GLPGU3ITJX1T8/twc+24+hr.png 02. Content Testing at Scale https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594934530512-FE7GZZV9LHMSHDXJCHCG/twc+custom.png 02. Content Testing at Scale https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594936009727-I8FWPP4IO1TX85SZD4N7/twc.png 02. Content Testing at Scale https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594936030883-K9O8PZU0E6SWO1AYX3BQ/alexa.png 02. Content Testing at Scale https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594936048689-QCKGQKRASN0KJA7NB746/big+sky.png 02. Content Testing at Scale https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594675258190-3W625T1QWBV0VWFA9XOS/Dc+3.png 02. Content Testing at Scale https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594675060607-EIT3N1K4ZBBFQMGILKSV/DC+1.png 02. Content Testing at Scale https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1594675172302-B0M3YNN8MUBMBG5RWH07/DC+2.png 02. Content Testing at Scale https://www.chadco.co/connect daily 0.75 2020-08-24 https://www.chadco.co/convo daily 0.75 2020-07-17 https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1595000198832-P1O881E270ZK67ZNM7HD/Screen%2BShot%2B2020-07-17%2Bat%2B10.32.59%2BAM.jpg Conversational Marketing https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1595000112783-EZ1KPPI5B321WSAF13WW/businessman-is-texting-on-smartphone-vector-illustration-vector-id1192503516.jpg Conversational Marketing https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1595000085173-PDEFYOL7QEHS0YF94T6V/social-media-networking-chatting-texting-communication-online-community-posts-comments-news-flat-vector-illustration.jpg Conversational Marketing https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1595006640231-YD5JYYEF5NH83DRUHT2X/WAds+-+chat+only.002.png Conversational Marketing https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1595008291023-A1LLMLLJ51B1ZI35QXU0/Screen%2BShot%2B2020-07-17%2Bat%2B9.49.15%2BAM.jpg Conversational Marketing - Meet Jonathan. 34 years-old and living with a partner. Someone in this stage of life might be thinking of putting down roots and might want to purchase a home. Mortgage lending can feel opaque, and, with all the factors that go into making the right decision and getting a good deal. Maybe, he feels informed or at least semi-informed about the process. Maybe, he wants to talk to someone to understand what he doesn’t know with out the pressure to apply immediately. The financial crisis of 2008 and the recent economic downtown with Covid give him pause. He’s suspicious of lenders who might not offer the best solution for him and his partner. https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1595008365498-8GQ2F37FBJJ3QQ2HLXRB/Screen%2BShot%2B2020-07-17%2Bat%2B9.49.47%2BAM.jpg Conversational Marketing - Meet Sophie. 27 years-old, and living it up in those first years out of school! She’s aware of all the adulting she does now. And, she thinks a lot about health, and how she’s always eating out. And, she’s concerned about the environment. Maybe she thinks it’s time to buy a grown-up car. Maybe she wants to get something she thinks she’ll really like and something that isn’t a gas guzzler that’s going to single-handedly destroy the planet. Sophie hasn’t done a ton of research, but she wants a way to better understand everything, without having to sort through a bunch of jargon. https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/1595011974788-UNF3AU9JKPFEUFKIRDUA/WAds+-+chat+only.003.png Conversational Marketing Considered here, another conversation carried on with the same user, across the purchase funnel. Here, the messages delivered in both and chat and audio. It’s not a stretch to think that this third representation could be driven by a video ad. https://www.chadco.co/google-ccai daily 0.75 2022-04-18 https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/9da844a9-289d-4bdc-8930-f6d4c57e6929/Screen+Shot+2022-04-18+at+2.27.31+PM.png Google CCAI - It starts with data. Conversational interfaces rely on language as input. If the bot can’t understand your users’ statements and process them accordingly, users will never experience the great thing you’ve designed. In the first stages of this project, I was able to oversee a team of linguists as they reviewed customer transcripts, identified head / contextual intents, and organized those intents into a useful taxonomy or Information Architecture for our bots. https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/5f1d7474-0dd2-4d9c-a7aa-7b6355eb92a2/Screen+Shot+2022-04-18+at+2.28.23+PM.png Google CCAI My team was able to create a system for data classification. Creating a information architecture or taxonomy defined as follows: a system of classification realizing a hierarchy of schema / instance relations. Ours is a verb based taxonomy that incorporates nouns to provide a uniform but readable schematic for undestanding broad topic areas (drivers), user or agent actions, and user or agent objectives. This approach became the Google CCAI standard for intent and TPU classification. In practice, this system saw our client realize a gain in F1 score from 0.56 to 0.83 for the chat bot within 2 sprints of launch https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/e837fdb1-b8c3-4bb3-b7ea-c61005ee7410/Screen+Shot+2022-04-18+at+2.57.13+PM.png Google CCAI - Bot design and development With taxonomy set, I led the team through the design and development of bot experiences. I facilitated the creation of design systems and oversaw bot development in Dialogflow CX. Our bots met or exceeded Google benchmarks of 30% containment for billing inquiries and ~50% containment for account management inquiries. https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/4e675e91-77c7-4bfa-94d8-e53c80cccbd8/disambiguation.jpg Google CCAI https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/a7547f61-20ef-4812-a76e-5e47742cd751/welcome.jpg Google CCAI https://images.squarespace-cdn.com/content/v1/5f05dd9e7002d43baebe1a2f/4517aefa-6d72-4ff7-bfa9-fbede39e181c/automation.jpg Google CCAI