Permission Not Required

Our last episode, at least for a while. Colleen accepted a full-time job and Joe's third kid is due in September, so we're wrapping up Permission Not Required with a good one.Joe went all-in on Ruby Native at RubyConf. No Hotwire Native leads this time. He ran a workshop, hosted office hours, and pitched in Ruby Central's new Ruby Runway showcase. Two people subscribed during the conference and six signups traced back to it. One's running Microsoft SAML, so he's charging a real business $300/year and has to fix that. He didn't go outside once in four days. The hallway track still won.The real ROI of a conference is relationships. You can't judge it by the signups that week. The payoff is the relationship that becomes a client or a business years later. Joe and Colleen met exactly once at RailsConf and here they are. Colleen recommends Scott Galloway's The Algebra of Wealth on relationships as a core life pillar, and her rule is simple: show up more than once.Colleen is obsessed with her self-hosted AI assistant. Three kids, seven calendars, an inbox from hell. She rebuilt her assistant on Hermes on a Hetzner box (goodbye OpenClaw) and runs three Telegram agents: a main VA, a health nag, and an email digest. It tracks her open loops and "mosquito tasks" and slots to-dos into the free blocks on her calendar. Security PSA: wiring an agent to your email opens a password-reset attack vector, so she hardened everything and kept email read-only.Parenting-meets-AI advice. Set up a shared kids@ email now, because you can't change it with a school district later. And Joe tried to let Claude read his consulting-lead emails through the Gmail MCP. Google shut the account down in four hours for looking like a bot.Then the news: Colleen's new role is AI enablement in-house, the same work she does now at a much bigger scale. Thank you to everyone who showed up and listened.Links:linkedin.com/in/[email protected]@masilotti.com

Joe and Colleen are back from the Fourth of July, Joe from a week on Lake Tahoe and Colleen from the all-out celebration her island throws every year. Then Colleen walks through a week of reflection: she wants to double down on her strengths instead of shoring up her weaknesses, and she's finally admitting she doesn't want to build an agency, even though she's confident she could grow one to a million dollars in five years. It just wouldn't get her any closer to what she actually wants.Joe is heads down for RubyConf, where he's running three Ruby Native sessions: a community hack day, a Ruby Runway finalist pitch, and a workshop. He's also wrestling with what Ruby Native even is. Is it a self-serve tool for solo developers at $299 a year, or the thing he uses to deliver higher-touch consulting? A cold lead just picked Ruby Native over a custom Hotwire Native build, turning a $25-30k, three-month project into a $15k, six-week one, and that captures how price-sensitive the whole market has gotten this year.They close on a pattern they keep seeing: everyone generates Claude documents, and then everyone feeds those documents back into Claude to summarize, until the actual thinking gets smashed through a sieve. Write it for a human, not for an AI. Joe also turned Fable 5 loose on the Ruby Native codebase for 30 minutes and got 111 findings back, a few genuinely useful and a few completely absurd.Chapters00:00 Vacation Stories02:58 Journey of Self-Discovery11:00 Agency vs. Startup18:00 RubyConf Preparation24:15 Pricing Strategy and Customer Segmentation26:01 Consulting vs. Self-Serve Product27:01 Adapting to Changing Consulting Landscape28:01 Price Sensitivity and Client Conversations30:14 Consulting Business Model and Future Plans32:18 AI and Human Behavior in Document Creation38:01 Outsourcing Thinking and Deep Reading App

Joe's premium custom builds aren't closing anymore. For the first time, clients are walking on price, reasoning they can just throw Claude tokens at it for a fraction of the cost. Ruby Native lets him deliver the same value at a much lower price point, but he's still wrestling with what that does to his business.Colleen, fresh off Legoland and a grown-up Vegas trip, pushes back: the value isn't dropping, the market is shifting, and the real problem is distribution.They dig into a funding conversation Joe has tomorrow, why money isn't a silver bullet, and the partnerships and marketing experiments that actually move the needle.Chapters00:00 Vacation Adventures07:05 Returning to Work16:50 Pricing and Value22:07 Ruby Native Market Position29:00 Distribution Strategy and Experiments36:03 Taking Investment and Marketing Experiments

Colleen gets her first case study back: a client's blog traffic jumped 86% from the SEO content pipeline she built. The catch is the maintenance. Every client sits on their own Hetzner box running OpenClaw, and it falls over every few weeks. Joe's pitch: stop fighting OpenClaw and make Rails the wrapper. An internal tool, single-tenant, add clients in code, and own the box.On Joe's side, there are now four Ruby Native apps in the App Store, all running the exact same code off one config file. He's reaching back out to past Rails clients with a clever offer: hand over your source for the weekend and get a working prototype on TestFlight by Monday. Then take the code and self-serve, or pay to get it shipped. He's weighing a Ruby Weekly sponsorship and going all in on Ruby Native at RubyConf with a workshop, the hack day, and a Ruby Runway finalist spot.They dig into why Google Ads aren't paying off in two to three months, why "go-to-market automations" beats inventing a market, and whether OpenAI ads are worth touching yet. Then Joe's confession: he leaned on AI to write, a long-time reader caught it, and he swore off AI for the actual words. The human post he wrote next hit the Hacker News front page before the admins buried it.Chapters00:00 Family and Summer Activities07:54 Product Development and OpenClaw Challenges13:05 Marketing Strategies for Ruby Native20:43 Focusing on Ruby Native at RubyConf26:39 Colleen's Consulting Approach and Content Automation36:12 Summer Mode and Work Schedules

Colleen had a weekend revelation: she doesn't love building products as much as she loves building businesses. The work doesn't matter, the industry doesn't matter, and tech might not even be the long-term home. Joe pushes back, then realizes she's been circling this for years.They float a partnership experiment where Colleen brings Joe qualified leads and he delivers the work, then dig into the harder problem underneath: how anyone builds trust at scale when cold outreach doesn't convert anymore and the real sales cycle is years long.Joe ships pingrb, an iOS and Android app for push notifications from Stripe, GitHub, Hatchbox, or any webhook (including Claude Code). He built the whole thing without opening Xcode or Android Studio, and it's the first Ruby Native app he's used as a customer instead of a maintainer.Also covered:Colleen's Google Ads MCP server is finally approved (took two YouTube walkthroughs)Google Play's wild "12 users for 14 continuous days" rule for personal accountsJoe on refusing to keep up with the AI shininess raceColleen's background coding agent on Hetzner that runs from SlackRuby Native's new screenshot pipeline: simulator captures, in-browser editor, one click to App Store ConnectThe funnel problem: five years of brand building, no lever to cash in onChapters:00:00 May Gray and false spring01:20 Try AdWizard MCP server approved by Google04:42 Google Play's 12-users-for-14-days rule08:36 "Maybe I don't actually love tech"17:24 Threads of this going back years19:39 Colleen finds the leads, Joe delivers the work22:22 Pull, not push: trust at scale26:25 Refusing to keep up with AI27:24 Background agents on Hetzner32:22 Ruby Native ships Android, plus PingRB35:45 The new screenshot pipeline38:00 Which marketing channel for Ruby Native?42:30 Five years of brand building. Now what?

Joe officially launched Ruby Native for iOS on his birthday with a 33% off discount code. The launch itself was quiet, but the bigger milestone landed the next morning: a developer in Turkey shipped their personal finance app to the App Store using Ruby Native, never opening Xcode, never writing a line of Swift. It's the first Ruby Native app in the store that Joe didn't build himself.Colleen pushes back on the marketing angle. The technical, craftsman framing only reaches a narrow slice of developers. The bigger TAM is indie devs trying to make money. Ruby Native unlocks B2C App Store distribution for Rails apps, and that's the pull. Joe sees how this could split the audience: hobbyists and prosumers on the $299 or $999 annual plan, businesses that need custom native code on his consulting side. The product becomes lead gen for the consulting, not a threat to it.Colleen also shipped a free Google Ads MCP server at tryadwizard.com. It's intentionally not a product. It's lead gen for her AI consulting. Next week's episode will be a live stream where Colleen sets up a real Ruby Native ad campaign with Joe, end to end.Joe wraps with a consulting update. Two inbound Mobile Playbook calls in the last week from longtime newsletter readers, plus a cold email reply that turned into a booked call. If all three close, a rough year turns into a normal one, just in time for baby number three in September.Chapters00:00 Introduction and Updates09:57 Target Market Considerations17:02 Partnership Opportunities23:24 Building and Using MCP Servers for Google Ads Management

Colleen goes to a Claude meetup in San Diego expecting 12 people. 80 show up. She wings a presentation and watches a winery owner and a lawyer demo working software they built with no code. The most interesting builders in the room aren't developers.Joe pitches Ruby Native for the Ruby Central pitch competition being held at RubyConf, forcing him to write a real business plan for the first time. He shares why signups are converting faster than anything he's launched before, and why the next positioning shift is away from "save developers time" and toward "get your Rails business into the app stores."Colleen reframes the insight: making someone money is more powerful than saving them time. Joe sees his whole pitch differently by the end of the conversation.They also talk about motion vs. action with back-to-back discovery calls, Colleen's first cold-lead proposal going out, Joe's LinkedIn content workflow built on top of his own writing archive, and why shipping a demo video in ScreenFlow beat a week of fighting Remotion.Chapters00:00 First World Problems and Lipstick Fiasco23:26 Navigating Client Calls and Offers28:28 The Value of Client Calls and Networking36:08 Video Content Creation and Launch Strategy43:44 Exploring Ad Creation and Strategy

Joe's Ruby Native launch video is stuck in Remotion hellClaude + Remotion produced something mediocre after many rounds. Colleen's verdict: commit to becoming a video creator, or pay a pro. Joe admits he has a mental block around spending business money, even though he sells consulting for a living.Joe migrated 7 apps to Hatchbox over the weekendOff Heroku, Render, and Fly onto a single $25 Hetzner box + $10 Hatchbox to save ~$1,500-2,000/year. Postgres → SQLite, Solid Queue, nightly S3 backups. Heroku's still the gold standard, Fly's dashboard is unusable, and every "Heroku but better" startup ends up being Hatchbox. The dream of building their own PaaS is officially dead (again).Colleen got her first cold-ish inbound for AI consultingBut the asks are vague: "I feel like I should use AI." Clients treat it like a magic lead-gen machine. Her fix: send tiered proposals ($5k / $20k / $50k) so the client picks their own budget and scope creep gets a natural guardrail.Breaking into non-tech industries (insurance, mortgage brokers)Huge opportunity, but enterprise procurement kills independent pitches. Joe floats: pitch a 6-month W2 embed to learn the institutional knowledge and build the system from the inside. Case study gold.LinkedIn 30-day challengeColleen posted 7 days straight, impressions up 400%. LinkedIn's feedback cycle is weird and slow (posts resurface for two weeks), but it's clearly shaping social proof on discovery calls. X has plateaued, Substack feels like "X with a smaller audience." Joe commits to daily LinkedIn starting April 3.Chapters00:00 Challenges of Video Creation07:58 Migration to Hatchbox and PaaS Limitations36:15 Social Media Engagement Strategies

Colleen is two months into her AI automation pivot and doing the work, but cold inbound leads haven't landed yet. She's going all-in on LinkedIn for 90 days and doing discovery calls in insurance and home services. Joe got two full-price annual Ruby Native subscribers over the weekend and is building Inertia support for a well-known client headed to TestFlight. They swap conference networking strategies, debate whether there's a magic industry to target, and vent about Anthropic killing subsidized tokens on third-party harnesses.Chapters00:00 3D Printing Adventures07:02 Returning from Vacation14:59 Networking at Conferences20:04 Industry Exploration27:06 Deterministic vs. Indeterministic AI37:28 Marketing Strategy and Audience Targeting43:15 Engagement and Intent Signals

Colleen is overwhelmed in a good way after saying yes to everything, and Joe pushes her to stop planning and just run ads for two weeks. Joe hits 1K ARR on Ruby Native but can't bring himself to launch it. Colleen pushes back on scope creep and they land on a two-phase launch plan. Plus, the irony of AI making both of them work more, not less.Chapters00:00 Life Updates and Overwhelm07:00 Exploring New Markets12:59 Balancing Work and Productivity21:57 Client Work and Side Projects29:48 Expanding Scope and Product Features36:21 Market Validation and Launch Strategy42:17 Planning the Product Launch47:25 Final Preparations and Launch Strategy