Support Lessons from Lyft, Square, and Google, with Beth Trame

Jenny Dempsey spent years leading support teams at tech startups before a layoff led to a two-year job hunt. A chance encounter with a junky old table in a friend's basement led to Jenny's furniture-flipping side business, and then her career-flipping podcast. Jenny and Mat get into caring intentionally, using AI to help you work out your values, and what to do when a friend is out of work.Some key moments(1:47) A layoff, a long job hunt, and a junky table in a friend's basement(11:54) Caring with intention, not caring about every freaking thing(16:22) "Just lay me off. Let me learn these lessons"(18:32) Five values on a Post-it note(21:25) AI isn't a therapist, but it can ask good questions(23:09) How to prepare while you're still comfortable(26:34) Jenny's Career Flipper podcast interviews(35:58) How to help a friend who's out of workToo long, didn't watch? Here's a concise versionDon't care less, care more deliberately. Caring is important, but save it for what really matters and try to let the other stuff slide.Get ready while you're on solid ground. Update your LinkedIn and build connections now, before you need them, and have something in your life beyond your job.Take your out-of-work friend to dinner. You don't need to send them job ads or ask how the search is going. Show up, keep it simple. Send them a funny dog video. Or a cat one. You know them best.

Should you trust a chatbot when it sometimes makes mistakes? Javier Silverio has spent eight years in the Help Scout support queue, and now builds the team's internal AI tools.He read Mat Patterson's Supportive Weekly newsletter on AI and trust, and they get into what it means to hand your data to a company when you can't check what happens to it, why customers have been so quick to forgive a wrong answer from a bot, and how trust tends to get built at the moment something breaks. Some key moments(3:27) What DJ requests and feature requests have in common(6:41) "Should I build it just because I can?"(7:47) The Monzo co-founder who records every conversation(12:16) Trust gets built when something goes wrong(21:21) Some people will wrestle with a bot, some just want the answer(23:34) A year spent teaching an AI tool to say "I don't know"LINKS FROM THIS EPISODEShow notes: https://helpscout.com/blog/javier-expected-complaints/Javier on LinkedIn: https://www.linkedin.com/in/javsilverioThe Tom Blomfield talk Javier mentions: https://www.youtube.com/watch?v=X_JsIHUfUjc&t=140sMat's article on trust, AI, and customer service: https://www.helpscout.com/blog/trust-ai-and-customer-service/The newsletter Javier reacted to: https://email.helpscout.com/our-new-insentient-overlordsLISTEN TO THE SUPPORTIVESpotify: https://open.spotify.com/show/61pw99zFj5MnP5I3L8S26dApple Podcasts: https://podcasts.apple.com/us/podcast/the-supportive-podcast/id1743286439Newsletter: https://www.helpscout.com/newsletter/supportive/The Supportive is Help Scout's podcast about customer service and the people who do it, hosted by Mat Patterson.#customerservice #AIchatbots #customersupport

Dave Berner is a software engineer who, as a two-time founder considers himself "almost unemployable". Dave chats with Mat about careers in an AI-disrupted world. Dave chats with Mat about using AI as a sounding board, keeping people in the loop, and how he thinks about career planning when so much of the future remains uncertain. Watch the full interview here or subscribe to The Supportive in your podcast player of choice.Some key moments(2:36) Using AI as a thinking partner and not a hype man(8:19) Is Dave disrupting or enabling? (11:15) Who's responsible when AI gets it wrong?(13:20) Doctors and their AI diagnosing problem(19:24) Why Dave stopped trying to pick one job title(23:22) Building your network before you need it(29:51) Embracing AI without losing yourselfToo long, didn’t watch? Here’s a concise version:Trust the person behind the AI- Dave says "You're basically paying for insurance" by having a trusted, informed person to verify and be responsible for a decision even when AI made it. AI as a sparring partner, not an oracle.- Conversational AI can help you figure out your own answers, rather than tell you what to think.Start asking questions now. - Find the people you trust and ask them what they've noticed about your strengths, and the areas that might need work. Visit https://helpscout.com/blog/almost-unemployable/ for all the links and notes.

Nicole Crayne, self-proclaimed "Chaos Wrangler" at Saffire, spent 12 years between support roles wrangling the chaos and the curiosity of kids as a stay-at-home parent. Now back in the corporate world, Nicole chats with Mat about what it takes to build curiosity into a team - trust, slack time, and a type of disciplined approach to asking "why".We also learn how her dad has sustained a decades-long plumbing business on a specific attitude towards service. See https://helpscout.com/blog/fix-your-plumbing/ for show notes and links.

Beth Trame has built support organizations from scratch multiple times, including at Lyft and Square, both businesses that dealt with steep volume increases and diverse customer bases. In this episode, Beth shares her experiences as we discuss customer feedback, criticism, and how support teams respond to them.Some key moments(0:19) Meet Beth Trame: Square, Lyft, and 10 years at Google(1:21) The newsletter: "Nobody Likes to Be Criticised"(5:45) The three-way relationship: driver, passenger, and the business(12:29) One-star reviews and reaching out to drivers directly(16:35) The A/B test that split CSAT into product vs. agent(25:28) Project Aristotle and psychological safety(35:21) Beth's advice: be a listener firstFor show notes and links visit https://helpscout.com/blog/customer-feedback-lessons-from-lyft-square-google/

Tom joined Help Scout's support team in 2026, but he's been using Help Scout for a decade. He chats to Mat about his role in improving Zapier's support team diversity, what actually changed to hit their goals, why an interview process that never changes isn't neutral, and how improving diversity also improved camaraderie.Key moments:(5:24) Tom's alottment in the UK, and companion planting as team building(7:26) Zapier's project to make their team reflect their wider country(13:33) Creating a more accessible and fair interview process(19:11) Culture fit is a signal, not a deciding factor(20:44) Camaraderie built by diversity(29:43) Everyone already has a hiring rubric, even if it's unwritten

Thinking about switching help desks? Help Scout support pro Chrissy Chavez tells Mat why the urge to burn it all down and start fresh is so strong — and why an audit should always come first.Chrissy is a senior technical support specialist based in Oahu, Hawaii. She joins Mat to react to his newsletter "Why You Shouldn't Switch to Help Scout" — and the conversation turns into a playbook for anyone eyeing a support tool switch: what to audit before you move, how half of "missing features" turn out to already exist, and how to walk customers through a big tool decision without scaring them off.TIMESTAMPS:5:48 Chrissy watches "Love It or List It" — "you're messy no matter where you live"7:56 Half the "missing features" people want are already there9:51 You might not have to switch tools at all15:32 Audit your current tool first — saved replies, stale docs, historical data that could cause AI hallucinations20:14 The emotional work of change: "Two weeks from now, we won't even remember it was different"LINKS FROM THIS EPISODE:- Find Chrissy on LinkedIn https://www.linkedin.com/in/chrissychavez- Read Mat's original newsletter https://email.helpscout.com/switching-tools - [Mat's 9-step guide to switching help desks](https://www.helpscout.com/helpu/switch-help-desk/).- Watch some Love It or List It https://www.youtube.com/watch?v=HOaEcx-DyPQ#CustomerSupport #HelpDesk #CustomerService

Pavel drove AI customer support adoption as head of support...and then he got laid off. Mat chats with Pavel about his 13 years in B2B SaaS customer experience about the way execs are thinking about support today.Plus: why customers still want a human to confirm an answer, even when the AI got it exactly right.For a full transcript, notes, and links, see https://www.helpscout.com/blog/inside-ai-customer-supportFind Pavel at: https://www.linkedin.com/in/pmalyshev/

Leslie O'Flahavan has spent 30 years teaching people to write well at work — including plenty of customer service teams. Leslie and Mat discuss what it takes to write concisely in support, and how to develop your team's concise writing skills.Listen out for Leslie's simple 10% word-count game, and the surprising shift she's made just this year in her approach to spelling and grammar.For full show notes see https://helpscout.com/blog/customer-service-writing

After 700+ live Help Scout classes, Alison Groves has figured out what makes customer education work — mistakes. Well, not exactly, but that's part of it. In this episode of The Supportive, Mat talks to Help Scout's Customer Education Lead, Alison, about what she's learned, how to get started, and why imperfection builds trust. WHAT WE COVER: - Why 90%+ of Alison's attendees stay for the full session - The three types of people who show up (and what each of them needs) - Why making mistakes live is a feature, not a bug - How to start even if you have zero experience - What happens when only 3 people show up vs. 120 - Whether AI will make live human teaching more or less valuable TIMESTAMPS: 0:00 Introduction 0:32 Meet Alison — 700+ classes and counting 4:11 What makes a great teacher (and a terrible one) 5:41 Getting over the fear of being wrong 8:09 Why people sign up for live classes 11:30 Live teaching vs. recorded demos 14:07 The value of imperfection 17:59 Advice for getting started 20:06 Your first step: just record yourself 22:14 Big class vs. small class — does it matter? 25:25 How to know if it's workingFor notes and links: https://helpscout.com/blog/teach-your-customers/