Building Repeatables in Claude: Skills, CLI vs MCP and Token Discipline | Go With The Flow

Claude Skills, CLI vs MCP and Token Discipline with Ritu Java | Seller Sessions

SEO Description

Ritu Java and Danny McMillan on building agentic skills, choosing CLI over MCP, plan mode discipline and the short window to ship before token costs reset.

Episode Summary

Week 4 of the month, Go With The Flow, and Ritu Java is back from her travels. The world has shipped fast since the last episode: Codex 5.5, Claude 4.7, an Amazon Ads MCP and a fresh round of panic over the rumoured removal of Claude Code from the $20 plan (it was a 2% AB test, not a rollout). Ritu and Danny use the noise to make a sharper point: this is the moment to stop chasing models and start building repeatable systems on the platform you have already chosen.

Ritu walks through the three eras of PPC Ninja's automation stack. Apps Script bulk file generators three years ago, Netlify hosted UI apps last year, and now agentic skills that her team chats with in plain English to produce upload ready Amazon bulk files. The same shift applies to data: BigQuery accessed through the Google Cloud CLI rather than through MCP, because CLI is leaner on tokens and works better when the job is heavy on data rather than tool surface. Danny mirrors the move with his event-ops CLI for WordPress, WooCommerce, Stripe and FooEvents reconciliation, and his four tier ExtractFlow cascade (HTTP, headless, stealth, agentic) that bypasses the limits of any single browser tool.

The second half is a discipline talk. Plan mode every time. Push back on the first plan because Claude over engineers by default. 30% of your time on workflow scaffolding so the other 70% can be real building. The 21 day Claude rule: when a shiny new tool fires the dopamine, wait 21 days before refactoring around it. Left brain tasks (counting, SQL, deterministic logic) belong in scripts. Right brain tasks (judgment, creativity, hypotheses) belong in the model. Mix them inside a single skill. Skills are micro pieces of your workflow, not magic, and Claude can write them for you from an existing SOP.

Key Topics

• The three eras of PPC Ninja automation: Apps Script, Netlify UI apps, agentic skills
• CLI vs MCP: when to choose each and why CLI is more token efficient for data heavy work
• Token economics, the rumoured $20 plan change and why it was a 2% AB test
• The short window before subsidised tokens get repriced
• Plan mode discipline and the "push back on plan one" rule
• Danny's 30 / 70 framework: workflow scaffolding vs building
• The 21 day Claude rule for resisting tool churn
• Left brain vs right brain task design inside a single skill
• The PPC Ninja "5 Whys" skill: deterministic SQL plus non deterministic hypotheses
• Claude.md, Gemini.md, Skills.yaml and the emerging Agents.md standard
• Skills for beginners: let Claude write them from your SOP
• Skill cascading: research, article, LinkedIn post, tweets, slide deck in one chain

Timestamps

• [00:01] Welcome back, Week 4 Go With The Flow, Ritu returns from travels
• [00:17] Codex 5.5, Claude 4.7 and the "no one is writing code anymore" reality
• [02:01] Ritu on the three eras of PPC Ninja automation
• [02:42] Era 1: Apps Script bulk file generators in Google Sheets
• [03:46] Era 2: Netlify hosted UI apps with input fields
• [04:48] Era 3: Agentic skills, the bulk file skill trained on Amazon templates
• [06:22] Cloud talking to BigQuery through the Google Cloud CLI
• [07:00] Danny: what is a CLI and why it matters for token use
• [08:00] Amazon Advertising MCP vs CLI based access to the same data
• [09:33] WordPress horrible to drive via MCP, easy via CLI
• [10:00] Danny's event-ops CLI: tickets, food tickets, WooCommerce, Stripe reconciliation
• [12:13] ExtractFlow four tier cascade: soft, medium, stealth, agentic
• [13:46] Why CLI for the heavy stuff, MCP for the soft touch
• [14:13] AWS CLI: chat to Claude, push HTML blog posts live in two minutes
• [15:33] The overwhelm problem and the 5,000costbehindthe5,000costbehindthe100 plan
• [17:35] The $20 plan rumour: it was a 2% AB test, not a rollout
• [19:38] Build repeatables, not one offs
• [20:38] Danny: pick a platform and stop chasing benchmarks
• [21:16] The 21 day Claude rule for new tools
• [22:16] Plan mode every time, push back on plan one, get the second plan
• [23:02] Why am I building it, who is it for, what am I building
• [23:30] The 30 / 70 split: workflow scaffolding vs real building
• [25:13] Why long six to fourteen hour Claude runs are usually inefficiency
• [27:12] Compounding 1% a day across a year
• [27:47] "I build the things that build things"
• [28:00] Architecture vs apps: filling the gaps between A and B
• [29:06] Left brain vs right brain task design
• [30:01] Why throwing 80/20 at a sales drop diagnosis fails
• [31:33] The PPC Ninja 5 Whys skill: deterministic plus non deterministic in one flow
• [34:32] Claude.md, Gemini.md, skills.yaml and the agents.md standard
• [40:53] Beginners: let Claude write the skill from your SOP, use the interview pattern
• [42:39] Skill cascading: URL to research to article to LinkedIn post to tweets to slides
• [44:42] Mixing deterministic and non deterministic inside a single skill
• [45:39] Wrap up, signal to noise, who is it for

Key Takeaways

1. Pick a platform and stop chasing models. A new model ships every week. Time spent benchmarking is time not building. Double down on Claude (or whichever you chose), use the 21 day rule, and let the ecosystem catch up to the shiny thing in your feed.
2. CLI for heavy work, MCP for soft touch. MCP loads tools and skills into context and burns tokens. CLI uses programs already on your machine. For data heavy jobs (BigQuery, AWS, WordPress at scale), CLI wins. For light cross app workflows, MCP is fine.
3. Build repeatables, not one offs. Subsidised tokens will not last. The 100planreportedlycostsAnthropic100planreportedlycostsAnthropic5,000 to serve. Spend the window building scaffolding that compounds, not 14 hour vibe coding runs.
4. Plan mode every time, then push back. Claude over engineers by default. Generate the plan, then say "you have over engineered this, although I want it elegant, go back and review." Plan two is the one you start from.
5. 30% on workflow, 70% on building. Each new dependency, MCP, skill or repo you add to your workflow compounds across every future project. Stop building only the apps. Build the things that build the apps.
6. Left brain in scripts, right brain in the model. Counting, SQL, deterministic logic belongs in Python the moment you can offload it. Save the model for hypotheses, judgment and creativity. The PPC Ninja 5 Whys skill mixes both inside one flow.
7. Skills are micro pieces, not magic. Take an SOP, ask Claude to interview you with decision panels, and let it write the skill. Then cascade skills together: URL to research to long form article to LinkedIn post to tweets to slide deck.

Notable Quotes

"Instead of doing one offs, it is time to build repeatables. The more people can learn that skill now, the better it will be, because a year from now you may not have access to the same tokens." Ritu Java

"If you see something and it looks sexy and it has sex and sizzle and your dopamine is screaming to go after it, wait 21 days. Either Claude will have it, or someone will have a repo, and you can combine it." Danny McMillan

"Always use plan mode. Never accept plan number one. Tell Claude: you have over engineered this, although I want it elegant, go back and review. Then start from plan two." Danny McMillan

"I build the things that build things. I build the scaffolding the team needs so they can build on top of it." Danny McMillan

"Spend 30% of your time on your workflow and 70% building. The 30% compounds across every project." Danny McMillan

"If we just hand six months of ad, organic, ranking and SQP data to Claude with no structure, it is going to mess up. It will give you an 80/20 you are not satisfied with, because it is not equipped to handle that volume without scaffolding." Ritu Java

"WordPress is horrible to work with through MCP. It falls over all the time. CLI can be amazing for certain things." Danny McMillan

Resources Mentioned

• PPC Ninja : Ritu's Amazon PPC software and agency, base for the BigQuery + CLI stack discussed
• Claude Code : Anthropic's CLI for Claude, the primary surface used in the episode
• Anthropic Claude : Claude 4.7 referenced as the current model
• OpenAI Codex : Codex 5.5 mentioned as the rival shipping fast
• Google Gemini CLI : Referenced as a sibling agent surface (Gemini.md)
• Google BigQuery : PPC Ninja's central data warehouse
• Google Cloud CLI (gcloud) : The CLI Claude uses to talk to BigQuery
• Amazon Advertising MCP : Amazon's official MCP server for ads data, referenced as the MCP comparison point
• AWS CLI : Used by Ritu to publish HTML blog posts to ppcninja.com from a Claude chat
• Netlify : Hosting layer for PPC Ninja's previous era of UI based apps
• WordPress and WooCommerce : Backbone of Danny's event-ops CLI
• FooEvents : Ticketing plugin that lives behind WooCommerce in the event-ops flow
• Stripe : Source of the card fee variation Danny reconciles via CLI
• ExtractFlow / CloudExtract : Danny's four tier extraction cascade (HTTP, headless, stealth, agentic). Open repo
• Playwright : The default browser automation tier inside ExtractFlow
• Agents.md : Emerging AI agnostic instruction file standard alongside Claude.md and Gemini.md
• Sequential Thinking MCP : The MCP Danny invokes when asking Claude to step through analysis

Hosts

Danny McMillan : Host of Seller Sessions, founder of DataBrill, building AI native tooling and CLI based workflows for Amazon sellers.

• Website: https://sellersessions.com
• LinkedIn: https://www.linkedin.com/in/dannymcmillan

Ritu Java : CEO and co founder of PPC Ninja, Amazon PPC software and agency. Specialises in automation, BigQuery pipelines and agentic workflow design.

• LinkedIn: https://ca.linkedin.com/in/ritujava
• Website: https://www.ppcninja.com

What's Next

• Next week: Ritu and Danny pick up routines and the new Claude scheduler.
• In 8 days: Seller Sessions Live 2026 in London on 9 May. Last week to lock in any final discounts.

About Seller Sessions

Seller Sessions is the leading podcast for serious Amazon sellers, hosted by Danny McMillan since 2017. Go With The Flow is the weekly automation strand where Danny and Ritu Java work through agentic flows, MCPs, CLIs and skills, in real time, on the same stack their teams ship every week.

Episode published: 1 May 2026 Series: Go With The Flow (Week 4 of the month) Keywords: claude skills, claude code, cli vs mcp, mcp model context protocol, claude 4.7, codex 5.5, amazon ppc automation, bigquery cli, agentic workflows, plan mode, token optimisation, claude.md, agents.md, ppc ninja, ritu java, seller sessions podcast, go with the flow

Why Your Amazon Dashboard Is Lying to You + Remotion & Voice Cloning Reality Check | Claude Sessions

Amazon Dashboard Brain, Remotion Video & ElevenLabs Voice Cloning | Claude Sessions

SEO Description

Shubhash Sharma on building a data brain behind your Amazon dashboard. Danny McMillan on Remotion video and ElevenLabs voice cloning realities.

Episode Summary

Week 3 of the month means Claude Sessions, and Danny McMillan and Shubhash Sharma are back with a double feature for Amazon and TikTok Shop sellers building their own AI tooling. Shubhash picks up from last episode's SP API and Ads API walkthrough with a hard lesson learned the wrong way: a polished dashboard wired straight into Amazon is a window with no room behind it. The numbers will lie, and you will not know when a feed silently dies.

He walks through the fix: a "brain" sitting between the data sources and the dashboard. Supabase as the long term store, pgvector for unstructured stuff like contracts and reviews, n8n as the orchestration layer. Six core domains every seller shares (orders, products, analytics, ads, finance, affiliates and creators) plus an optional documents layer. He closes with a dual write migration pattern so you can flip between old and new without taking the business offline.

Then Danny turns to video and voice. Remotion looks like toy town out of the box, but with the right plugins (motion blur, transitions, captions, shapes, fonts, rendering) and Claude doing the orchestration, it becomes a serious production tool that can pull in your footage, branding and design system. On the voice side, he has tested VoiceBox and F5TTS and come back to ElevenLabs Multilingual v2 at £22 a month. The missing gap everywhere is cadence. He also names the deeper bet: as the market floods with AI generated content, authentic voice becomes the differentiator that cannot be cloned.

Key Topics

• Why dashboards lie when wired straight into Amazon, TikTok and Shopify
• The "brain" pattern: Supabase, pgvector and n8n as a centralised data layer
• The six core data domains every seller needs (plus a 7th for documents)
• Dual write migration so the old system and brain run in parallel
• Remotion as a code based video tool, and what it needs to stop looking toy town
• The four layer creative workflow: brief, story skeleton, treatment, scene by scene
• ElevenLabs vs VoiceBox vs F5TTS for voice cloning your own voice
• Why cadence is the last hard problem in synthetic voice
• The authenticity premium in an AI flooded market

Timestamps

• [00:00] Intro and welcome back to Claude Sessions
• [00:34] Shubhash kicks off: where to put the data you pulled last week
• [01:04] "Your dashboard is lying to you" and the polished dashboard pitfall
• [02:32] Dashboard is a window. The brain is the room behind it
• [04:54] Tech stack: Supabase (Postgres), pgvector, n8n
• [05:54] The six fundamental data domains
• [06:26] Orders, products, analytics, ads, finance, affiliates and creators
• [08:30] The optional 7th layer: unstructured documents via pgvector
• [09:44] Dual write pattern for safe migration
• [10:48] Three takeaways: audit, list domains, build one table at a time
• [12:28] Danny on Remotion: code based video and why it is toy town out of the box
• [13:51] What is missing: motion blur, transitions, captions, shapes, beat detection
• [14:54] The 80+ plugin packages that turn it into a real tool
• [16:56] Pulling in footage, logos, design systems and free music from Pixabay
• [18:30] The 4 layer creative workflow: brief, story skeleton, treatment, scenes
• [21:15] Voiceovers: ElevenLabs Pro setup and why the £22 is worth it
• [22:12] VoiceBox and F5TTS field test: garbage and 5 rounds of tuning later
• [23:22] Why cadence is the hardest thing for AI voice to fake
• [25:42] How much reference audio you actually need (30 min min, 2 hours ideal)
• [27:25] ElevenLabs UI parameters: speed, stability, similarity, exaggeration
• [28:52] The authenticity premium when the market floods with AI
• [30:30] Key takeaways, ElevenLabs API usage and locking in your voice once
• [34:24] Aside: "insane" and "most" as the new AI tells
• [36:31] SSL 2026 wrap, 18 days out, Ritu returns next week with Japan

Key Takeaways

1. Build a brain, not just a window. A dashboard wired straight to Amazon, TikTok or Shopify has no memory. When a feed silently fails, the dashboard happily lies. Sit a Supabase + pgvector + n8n layer in between, and your dashboard becomes a view on top of a real source of truth.
2. Six domains cover almost every seller. Orders, products, analytics, ads, finance, and affiliates / creators. Map every place each one currently lives, then consolidate one domain at a time. Start with one table (orders) and let Claude do the heavy lifting.
3. Use dual write when migrating. Write to the old store and the new brain in parallel for a week. Compare. Flip the dashboard's read side via a feature flag. If something breaks, flip back. Zero downtime, zero fear.
4. Remotion is a system, not a tool. Out of the box it is bare. Add the plugins (motion blur, transitions, captions, fonts, rendering), bring your own footage and design system, and let Claude orchestrate the four layer workflow: brief, story skeleton, treatment, scene by scene.
5. ElevenLabs Multilingual v2 still wins for voice cloning. VoiceBox and F5TTS were not close. Pay the £22, use Model 2, feed it 30 minutes minimum (2 hours ideal) of clean reference audio, and lock the setup in once.
6. Cadence is the last mile. AI can match tone and timbre. It still cannot match the rises, falls and micro pauses that make a sentence sound like you. Use scripts split into short paragraphs, generate three variants, and tune the language you use to talk to Claude until the cadence lands.
7. Authenticity becomes the moat. As written, visual and audio AI floods every channel, the brand voice that is unmistakably human becomes the differentiator. Do not give that away to save 22 dollars a month on a podcast.

Notable Quotes

"Dashboard is a window. We need a room behind the window. So the brain is going to be the room behind this window." Shubhash Sharma

"If any of our SaaS went offline tomorrow, will our business still have its memory? The answer is no, because we haven't stored it. All we have is rented attention." Shubhash Sharma

"When you migrate to your brain, don't rip out your old system. Use dual write. Run them in parallel for a week. If something breaks, flip it back. Zero downtime, zero fear." Shubhash Sharma

"Remotion out of the box isn't great. It's almost like building some slides, just one step up. You have to build it as a system of what you need." Danny McMillan

"The hardest part for AI to represent is cadence. It can get the tone of your voice. That's the easy bit. But the speed and the up and down of how you talk, that's where these models still fail." Danny McMillan

"In our rush to use AI, you've got to remember the market floods with it. When everything sounds like AI, the only thing left is the authentic voice for your brand." Danny McMillan

Resources Mentioned

• Supabase : Postgres backend used as the long term data store for the seller "brain"
• pgvector : Postgres extension for semantic search over unstructured data (contracts, reviews, supplier emails)
• n8n : Orchestration layer for scheduled pulls and cron jobs with a UI
• Amazon Selling Partner API (SP API) : Source for orders, inventory and finance data (covered in last episode)
• Amazon Ads API : Source for ad spend, campaign and keyword data
• Remotion : Code based, React powered video creation framework
• ElevenLabs : Voice cloning and text to speech. Model used: Multilingual v2 (Pro plan, £22 / month)
• F5 TTS : Open source text to speech model tested for voice cloning
• VoiceBox by Jamie Pine : GitHub voice cloning desktop app tested by Danny
• Pixabay : Free music and sound effects used inside the Remotion workflow
• Loom : Source of clean voice reference audio if you record team walkthroughs
• Seller Sessions Live 2026 : Conference 9 May 2026, 18 days out at recording

Hosts

Danny McMillan : Host of Seller Sessions and Claude Sessions, founder of DataBrill, building AI native tooling for Amazon sellers.

• Website: https://sellersessions.com
• LinkedIn: https://www.linkedin.com/in/dannymcmillan

Shubhash Sharma : Engineer building data infrastructure for Amazon and TikTok Shop sellers. Returning Claude Sessions co host.

What's Next

• Next week: Ritu returns from Japan with three subjects covered in this month's rotation.
• In 18 days: Seller Sessions Live 2026 in London on 9 May. Modular format, new venue confirmed.

About Seller Sessions

Seller Sessions is the leading podcast for serious Amazon sellers, hosted by Danny McMillan since 2017. Claude Sessions is the AI focused monthly strand where Danny and rotating co hosts work through the practical wins, false starts and engineering reality of building with Claude, MCPs and the wider AI stack inside real seller businesses.

In this Conversion Monthly, Danny McMillan is joined by Dorian and Matt Kostan (no Sim this episode — he's on holiday) for a live, practical session on building brand-quality design systems fast and for free.

Dorian opens with a tight crash course in the three design fundamentals that separate professional Amazon listings from amateur ones: font pairing, grid and layout, and colour theory. He then demos Google Stitch live, building a full design system from a wooden utensil listing in real time. Danny shows a more automated route — using Perplexity to control Stitch autonomously and generate a complete brand kit from just a product title, bullet points, and a reference image. Matt rounds it off with a live Product Pinion split test of the new designs against the original listing — and the results deliver the session's sharpest lesson.

The big takeaway: pretty is not enough. Information + design working together is what converts.

Key Topics

• Google Stitch for brand design — Free AI design tool that generates full brand guidelines, font pairings, and mockups from reference images and prompts
• 3 design fundamentals every seller should know — Font pairing, grid and layout, colour theory with a contrasting action colour
• Perplexity + Stitch autonomous workflow — Danny demos letting Perplexity control Stitch end-to-end with zero manual input to generate a full brand kit
• Coolers.co — Free colour palette tool with a visualiser and AI colour bot (Matt)
• UX and design laws applied to Amazon — Miller's Law, Fitts' Law, Jacob's Law, Occam's Razor translated into listing and brand site decisions
• Product Pinion live split test — New designed variants vs the original listing, with real shopper results in under 10 minutes
• Live test result — The original information-heavy image outperformed the prettier redesigns early on; lesson: strip information at your peril

Timestamps

• [00:00] Intro — Danny opens, Sim is out, format overview
• [00:48] Dorian: Why most Amazon listings lack design consistency
• [02:00] The 3 design principles: font pairing, grid/layout, colour theory
• [04:30] Font pairing explained — serif vs sans-serif, how world-class brands use them
• [07:00] Colour theory — complementary colours plus one contrasting action colour
• [08:30] Live Google Stitch demo — wooden utensil set, design system generated from brand brief + images
• [10:00] Stitch output: colour palette, font pairings, layout mockups
• [12:17] Matt: brand guidelines used to cost $1,000+ — now free in Stitch
• [13:00] Dorian: live Figma iteration — cleaning up the infographic using new design system fonts
• [17:00] Matt: information hierarchy lesson — measurements vs benefits on infographics
• [19:30] Dorian: "mouse text" and anchoring — what to leave in, what to strip out
• [20:33] Matt: Coolers.co overview — free colour palette generator and visualiser
• [22:00] Matt: UX/UI design principles applied to Product Pinion and Amazon listings
• [25:12] Danny: Perplexity + Stitch autonomous brand kit demo — Z Kitchen brand from scratch
• [27:00] Z Kitchen outputs: design system, A+ content, infographic, lifestyle mockups, packaging concepts
• [31:00] How to iterate inside Stitch — refine vs reimagine, varying only specific elements, up to 5 variants
• [36:00] Danny: UX design laws — Miller's Law, Fitts' Law, Jacob's Law, Occam's Razor
• [40:00] Danny: Typography slides — spacing systems, layout balance, font families
• [43:32] Dorian: reveals three redesigned variants ready for split test
• [44:35] Matt: launches live Product Pinion test — 50 shoppers, cooking category targeting
• [47:33] Live results coming in — original listing leading over new designs
• [48:00] Dorian: "pretty is one thing but the information has to be there"
• [49:00] Danny: design and information are two separate layers — both are required
• [51:30] Product Pinion API + Claude integration teaser
• [52:36] Final results and wrap-up — test completed in ~10 minutes with 50 real shoppers
• [53:44] Closing thoughts and Seller Sessions Live preview (26 days out)

Key Takeaways

1. Three principles separate professional listings from amateur ones — font pairing (serif + sans-serif), grid and layout (hierarchy: 1, 2, 3), and colour (complementary base + one contrasting action colour).
2. Google Stitch is the best free tool right now for design mockups — unlike image generators (Gemini, GPT), Stitch understands design principles and generates layout-aware mockups you can iterate on.
3. Pretty does not convert on its own — the live test showed the original, information-heavy image outperforming the cleaner redesigns early. Design is a layer on top of strong product information, not a replacement for it.
4. Perplexity can run Stitch autonomously — paste a product title, bullet points, and a reference image; let it loop through Stitch without touching anything; come back to a full brand kit.
5. You can test design variations with 50 real shoppers in under 10 minutes — Product Pinion lets you run image split tests with category-targeted shoppers, get qualitative feedback, and iterate the same day.
6. Nano Banana outputs in Stitch cannot be regenerated — switch to one of the standard models if you need variation or refinement controls.
7. AI gets you to the concept stage fast — use Stitch to generate the direction, then hand to a designer for finishing. Revision cycles and meetings shrink dramatically.

Notable Quotes

"If everything is important, nothing really is." — Dorian

"The hardest thing is to make something simple, elegant, and something that people get instantly." — Dorian

"Pretty is one thing, but the information has to be there. I didn't put the information there — and it's not doing well." — Dorian (on live split test results)

"Most people don't necessarily know good design, but they know what they like. It's more of a feel — they go, that looks a bit cheap, or that looks really good." — Danny McMillan

"It's never been easier and faster to become a world-class brand on design. Plug in your details, get a design guide going, and you can really up your brand in a very short period of time." — Matt Kostan

"The breakout brands from the Amazon community — we haven't had enough of them crossing over. Now that gap's closed." — Danny McMillan

Resources Mentioned

• Google Stitch — Free AI design tool; generates brand guidelines, font pairings, mockups, A+ content concepts, and layout variations. Up to 3,000 generations per day (free)
• Figma — Design tool used by Dorian to pull Stitch outputs and refine layouts manually
• Adobe Color (color.adobe.com) — Colour palette exploration and complementary colour tool; used in the live demo for the wood/blue beach-forest palette
• Coolers.co — Free colour palette generator with AI colour bot and real-world visualiser
• Pinterest — Recommended for browsing font pairing inspiration
• Nano Banana 2 — Image generation model available inside Stitch; note: regeneration/variation controls don't work on Nano Banana outputs
• Perplexity — Used to autonomously control Google Stitch via browser automation, building a full brand kit end-to-end from a single prompt
• Product Pinion — Consumer research and split testing tool by Matt Kostan; image tests with real shoppers, category targeting, results in minutes. Product Pinion API + Claude integration in development.

Guest Info

Dorian — Design and conversion specialist, Seller Sessions Conversion Monthly co-host

Matt Kostan — Founder of Product Pinion, consumer research and split testing for Amazon sellers

In this Broad Match Show, Danny McMillan and Adam Heist cover two of the most practical AI frontiers for Amazon sellers right now: getting direct API access to your Seller Central data and building a fully automated design workflow from inspiration through to live assets.

Adam breaks down how he connected Amazon's SP API and Ads API to an AWS database and wired Claude Code directly to it — giving him real-time, queryable access to years of business data across any metric. No developer required. Danny walks through his 8-step system that takes a seller from a TikTok scroll to a finished, conversion-tested design with brand consistency baked in.

Both share hard-won lessons on where AI gets you (the 70–85% mark) and where the human still needs to step in — plus a candid look at what's changing at Seller Sessions Live on May 9th.

Key Topics

• Amazon API data pipeline — SP API + Ads API → AWS database → Claude Code for real-time analysis
• 8-step AI design workflow — Inspiration capture, memory/photo brain, brand system, mood board, asset generation, build, and quality gate
• CLI vs MCP — Why CLIs are becoming the cleaner integration path for tools like Google Workspace
• Seller Sessions Live (May 9th) — New modular format, no sponsors, £5,000 fine system for service providers pitching
• Health check-in — Adam on fitness goals; Danny on resolving a high ferritin (iron overload) diagnosis

Timestamps

• [00:00] Welcome and introductions
• [01:10] Adam: Getting Amazon SP API and Ads API access as an individual brand
• [05:00] Storing API data in AWS and connecting it to Claude Code
• [07:30] Building custom dashboards and software from your own data
• [09:00] How to approach it if you're not technical — think first, screenshot issues, let Claude walk you through
• [12:25] Danny: 8-step AI design workflow overview
• [13:30] Step 1 — Inspiration capture from TikTok, YouTube, social reels
• [14:20] Steps 2–3 — Memory/photo brain + design system (52 world-leading brands baked in)
• [15:30] Steps 4–5 — TLDraw mood board + asset generation (Nano Banana 2, Gemini, Remotion)
• [17:50] Steps 6–7 — Build stage (React, Tailwind, ShadCN, Netlify deploy)
• [18:30] Step 8 — Quality gate (216-feature scoring: UX heuristics, typography, psychology)
• [19:30] Google Stitch + Perplexity demo: full brand system from a product title + screenshot
• [23:12] Adam: the 70–85% rule and how to think about AI-assisted design cycles
• [27:35] Danny: Google Workspace CLI for email — running launches under 3,000 contacts
• [29:26] Health updates — Adam on fitness; Danny on ferritin/iron overload and phlebotomy sessions
• [35:45] Seller Sessions Live May 9th — format, venue (inside a church), evening networking
• [41:49] The £5,000 fine system for service providers pitching at the event
• [43:01] Wrap-up

Key Takeaways

1. You can get Amazon API access as an individual brand — no developer credentials needed. SP API goes back 720 days; Ads API covers 60 days. Approval takes 1–2 days.
2. AWS as a data warehouse for Amazon data — pipe the API into AWS, connect Claude Code to it, and query anything: anomalies, stock-outs, week-on-week comparisons, year-over-year trends.
3. The non-technical workflow is: think → verbalize → screenshot issues → let Claude solve — you don't need to understand the infrastructure, just be clear on what you want to achieve.
4. AI gets you to 70–85% fast — bring in your designer or team at stage 4, not stage 0. Cycle times drop from 6 weeks to 1 week.
5. CLIs beat MCPs for tool integrations where available — less token overhead, fewer config issues, more cohesive experience in Claude Code.
6. Google Workspace CLI can replace Mailchimp/Klaviyo for small lists — Gmail allows up to 3,000 sends per day; viable for product brand launches under that threshold.
7. Seller Sessions Live is now sponsor-free and profitable on ticket revenue alone — the event model is shifting away from conference-style sponsorship dependency.

Notable Quotes

"Getting the actual real-time API data access has been just another level completely." — Adam Heist

"The original thought is: I need to get API access and I need to connect that to Claude. That's my thinking. And then you literally just verbalize that and use screenshots as you get stuck." — Adam Heist

"AI gets you to the finish line faster across way more dimensions, so instead of doing 600 things in a year, you're doing 2,000." — Adam Heist

"We live in a time whereby execution in a way is taken care of by AI. Where we're needed is on the vision — do we build this or don't we build it?" — Danny McMillan

"Know with AI it's dumb unless you give it a brain." — Danny McMillan

Resources Mentioned

• Amazon SP API — Business reports, inventory, listings, SQP data; up to 720 days historical
• Amazon Ads API — Ad performance data; 60-day lookback
• AWS (Amazon Web Services) — Cloud database for storing API data; connects to Claude Code via MCP or CLI
• Claude Code — AI coding assistant used to build the data pipeline and dashboards
• Google Stitch — Free UI design tool; used to generate brand systems from a product image + title
• Perplexity — Combined with Stitch to generate full design systems from Amazon listings
• Nano Banana 2 — Image generation tool controlled via Claude; used in Danny's asset generation step
• Gemini — Used with reference images for asset generation
• Remotion — Video generation component in Danny's design workflow
• TLDraw — Collaborative whiteboard/mood board tool; integrated with Claude for live-updating design boards
• React / Tailwind / ShadCN — Front-end stack used in the build step of Danny's workflow
• Netlify — Deployment target for the build step
• 21st Century Dev / ShadCN MCPs — Component library MCPs used in the build stage
• Google Workspace CLI — Cleaner alternative to Gmail MCP for read+write workflows in Claude Code
• Playwright / Fetch MCP — Browser automation tools; Danny built a 4-stage cascade scraper for Amazon

About the Show

The Broad Match Show is a monthly format on Seller Sessions, hosted by Danny McMillan and Adam Heist. It covers the cutting edge of AI tools, Amazon strategy, and brand building — first Tuesday of every month.

Seller Sessions is one of the longest-running Amazon seller podcasts, hosted by Danny McMillan. Known for deep-dives into conversion, data, and the practical application of AI for e-commerce brands.

Welcome back to Conversion Monthly! Danny McMillan is joined by the full team — Sim, Matt Kostan, and Dorian — plus special guest Anna, who launched a reusable microfiber cleaning cloth roll on Amazon UK.

Anna's product is clever — an eco-friendly alternative to paper towels that's washable and comes in multiple colors. But the listing isn't converting. The team digs into exactly why and what to fix.

What's Covered in This Episode:

• Main image problems — Why Anna's current main image fails to communicate what the product actually is, and how showing roll thickness, sheet count, and color options can dramatically increase click-through rate
• Baseline click share testing — Matt runs a 100-person UK shopper poll revealing Anna's listing captures only 10% of clicks against competitors
• The power of color — Shoppers gravitate toward listings showing multiple color options, and the team discusses how bolder, richer colors (not pastels) pop on search results
• Secondary image ordering — Dorian's favorite test reveals that simply reordering existing images based on what shoppers care about most (reusability and washability) can boost conversions without changing a single image
• Title strategy — Why keyword stuffing kills readability and how "speed bump keywords" like "washable" or "tear-away" differentiate you from competitors
• Cost-saving as a conversion lever — Framing the product as a money-saver versus annual paper towel spend gives instant price justification
• Before-and-after imagery — "Show me, don't tell me" — why context-driven images outperform generic product shots
• German marketplace analysis — What the German Amazon sellers are doing right with richer colors, dynamic angles, and clean layouts
• New main image results — After updating the main image with Dorian's mockup, click share jumped 60-70% in testing

Key Takeaways:

• If shoppers can't tell what your product is from the main image, nothing else matters
• The order of your secondary images matters as much as the images themselves — answer objections early
• Titles need to balance keyword volume with readability — a title nobody understands won't convert regardless of search volume
• Look at what competitors in other marketplaces (especially Germany) are doing for image inspiration
• Test before you invest in final designs — quick mockups and polling save time and money

Connect with the Team:

• Matt Kostan — [email protected]
• Sim — LinkedIn (also recruiting brand managers)
• Dorian — LinkedIn
• Anna — LinkedIn

Danny McMillan and Shubhash Sharma are back with another Claude Sessions episode covering both the back end and front end of building your Amazon business infrastructure with AI.

Shubhash walks through exactly how to register for Amazon's Seller Partner API — your free, direct access to your own sales, inventory, pricing, and order data — no third-party subscriptions required. Danny then breaks down the 165-feature design system he built to eliminate AI slop from websites, landing pages, and app interfaces.

Part 1: Amazon SP-API Setup (Shubhash)

• What SP-API is — Amazon giving you a key to your own data warehouse: live inventory, real-time orders, pricing, catalog data, and sales reports
• 5-step registration process — Register as developer, create an app, select permissions, self-authorize, and connect to Claude Code to build dashboards
• Common rejection reasons — Usually a missed checkbox or vague answer about data usage. Keep answers focused on personal brand development and safe data storage
• Advertising API is separate — Different credentials, different registration, different refresh token. You cannot reuse SP-API tokens for ads
• What you can build once connected — Custom dashboards, forecasting engines, inventory alerts, automated reporting — all built by Claude Code without knowing Python
• Danny's guardrails — Hire a $50 Upwork specialist to help with paperwork submission, keep them on retainer for when APIs go down (especially Q4, Black Friday, Prime Day)

Part 2: The 165-Feature Design System (Danny)

• The AI slop problem — Default fonts (Roboto, Arial), purple-blue gradients, three-column card layouts, floating animated orbs, oversized border radius — all telltale signs of generic AI output
• 15 anti-patterns cataloged — The system actively fights against common AI design defaults
• Four-phase pipeline — Decide, Design, Build, Refine — with 15 databases and components extracted from 11 repos
• Gap analysis scoring — Rates output out of 60 points. Seller Sessions Live went from 33 to 50; Databrill went from 48 to 55
• Psychology of design baked in — Hick's Law (limit choices to 5-7), Miller's Law (chunk information in groups), Jacob's Law — all running automatically in the background
• "Pretty doesn't convert" is a cop-out — Apple, Ralph Lauren, Sony all prove that quality design builds trust. The real issue was budget — now AI removes that barrier
• Design is about subtraction — Cut 69% of animations in one project. Overcooking destroys user experience
• 25 quality gate techniques — Color tokens, typography rules, contrast ratios, accessibility (100+ rules), spacing, and composition patterns
• Claude Loom workflow — Record feedback via Cmd+Shift+L, Claude extracts screenshots and browser URLs, and the system pushes back if changes violate the design system

Key Takeaways:

• SP-API is free and gives you direct access to your Amazon data — do it tonight
• The Advertising API requires a completely separate registration process
• Have a backup developer on standby for API downtime, especially during peak sales periods
• AI-generated interfaces all look the same because they default to the same fonts, colors, and layouts
• A design system isn't about making things pretty — it's about trust, conversion, and consistent user experience across all devices
• Before your customer reads a single word of copy, your page load time and visual quality have already made an impression

Coming Next:

Shubhash experiments with running AI models locally on an old MacBook using Ollama — cutting token costs to zero.

Connect:

• Shubhash Sharma — LinkedIn
• Danny McMillan — sellersessions.com

In this third installment of Claude Sessions, Danny is joined by Subash from Not A Square, who helps e-commerce brands scaling past seven figures implement AI without scaling headcount. Subash walks through real client case studies -- including a TikTok brand that boosted its customer satisfaction score from 4.2 to 4.5 in four weeks using a customer support agent built in Claude.

Danny then breaks down OpenClaw, the open-source personal AI agent that exploded in popularity, explains why he chose not to use it despite the temptation, and reveals Claude Flow -- his custom operating system built inside Claude Code with 11 engines, 300+ features, and a persistent memory layer powered by ChromaDB. The episode drives home one core message: document your operations first, pick one platform, go deep, and stop chasing every new tool.

Key Topics

• Documenting operations before automation -- Why you cannot automate what is not documented
• TikTok customer support case study -- Building an AI agent that raised satisfaction scores in four weeks
• OpenClaw overview and security risks -- What it does, why it blew up, and why Danny built his own alternative
• Claude Flow -- Danny's custom operating system inside Claude Code with persistent memory
• The amnesia loop -- How context loss between sessions kills productivity and how ChromaDB solves it
• Pixel-less environment -- The shift from structured prompts to contextual AI interaction
• Go deep on one platform -- Why chasing multiple AI tools guarantees you build nothing

Timestamps

• [00:00] Introduction -- Claude Sessions Week 3, delayed from the road
• [01:03] Subash introduces himself and Not A Square
• [02:01] Overview of three client projects and the problem founders face
• [04:30] Why operational truth is the moat in AI commerce
• [06:48] Three pillars: reduce costs, better governance, scale without headcount
• [07:30] TikTok case study -- customer support agent boosting store score from 4.2 to 4.5
• [09:04] OpenClaw -- history, capabilities, and the security nightmare
• [15:30] Six core capabilities of OpenClaw (local-first, universal messaging, persistent memory, browser automation, system access, self-extending skills)
• [18:00] Why OpenClaw matters -- moving from dumb LLMs to personal AI agents
• [20:00] Security trade-offs -- 1.5M API keys exposed, malware in skills, Cisco tests
• [22:00] Claude Flow -- Danny's 11-engine operating system built inside Claude Code
• [24:26] The amnesia loop -- how sessions lose context and how ChromaDB fixes it
• [28:19] Why Claude MD, agents, and skills are not enough without hooks and triggers
• [32:40] Go deep on one platform -- stop chasing every new tool
• [35:35] Subash on helping sellers adopt Claude Code fundamentals (Claude MD, skills)
• [39:51] Wrap-up and contact info

Key Takeaways

1. Document before you automate -- If your business operations live in the founder's head and not on paper, any AI tool will amplify the chaos rather than fix it.
2. Operational truth is the moat -- Clean inventory, accurate catalogs, honest cashflow reporting. Get these right before touching AI.
3. One AI agent moved the needle -- A single customer support agent on TikTok raised a brand's satisfaction score from 4.2 to 4.5 in four weeks, directly improving store visibility.
4. Persistent memory changes everything -- ChromaDB captures decisions, patterns, and project context across sessions so Claude compounds in usefulness over time (zero entries in session one, 1,700+ by session 25).
5. Scaffolding beats raw building -- Danny's Claude Flow system means a project that took five days six months ago now takes 40 minutes. The investment in infrastructure pays exponential returns.
6. OpenClaw is proof of concept, not production-ready -- Broad permissions, prompt injection vulnerabilities, exposed API keys. Wait for the open-source community to patch the holes before diving in.
7. Pick one platform and go all the way in -- Chasing multiple AI tools means you learn none of them deeply and build nothing of value.

In this Conversion Monthly episode, the team takes on a real-world Amazon fashion listing -- a ladies jumper from Adam Jagger's clothing brand. Adam launched this knitwear product last year with strong initial sales, but the relaunch has stalled despite refreshed images and PPC data.

Matt Kostan shares consumer feedback from Product Opinion videos revealing that shoppers found the secondary images too text-heavy and "fast fashion" looking, with inconsistent color rendering across photos. Dorian then presents a complete visual overhaul inspired by Zara's photography style -- low-angle shots, model-camera interaction, and stripped-back secondary images that let the product speak for itself.

The result: a 45% improvement in click share during simulated testing, jumping from 11% to 16% against competitors. The episode is packed with actionable advice on fashion-specific listing optimization, the power of "less is more" in secondary images, and why pre-launch polling is essential in design-led categories.

Key Topics

• Live listing teardown -- Adam Jagger's ladies jumper analyzed by the full Conversion Monthly panel
• Consumer video feedback -- What real shoppers said about the listing (color inconsistency, text-heavy images, fast fashion feel)
• Main image testing -- Baseline vs. Zara-inspired low-angle photography concepts
• Secondary image overhaul -- Stripping back text and adopting a premium, warm aesthetic
• Pre-launch polling -- Why design-led fashion products should be tested before manufacturing
• Selling through the female lens -- Understanding emotional and aspirational buying in women's fashion

Timestamps

• [00:00] Introduction and welcome to Adam Jagger
• [01:24] Adam explains the product -- ladies jumper relaunch that stalled
• [02:28] Category rules -- how much creative freedom do clothing listings have?
• [03:45] Matt shares Product Opinion video feedback from real shoppers
• [07:25] Dorian's approach -- studying Zara and H&M for photography inspiration
• [10:30] The "less is more" philosophy for fashion secondary images
• [14:02] Matt reveals test results -- 11% to 16% click share improvement
• [19:22] Sim's take on AI-generated images and warmth in the image stack
• [20:39] AI-generated video concepts for sponsored brand ads
• [22:50] Adam's reaction and takeaways
• [26:17] Danny's deep dive -- cognitive overload, decision paralysis, and the female lens
• [37:31] Adam's final thoughts and next steps
• [38:30] Sim on pre-launch polling for design-led niches
• [40:47] Final roundup and upcoming Seller Sessions event announcement

Key Takeaways

1. Strip back your fashion images -- Text-heavy, icon-filled secondary images can make clothing look like fast fashion. Clean, warm, product-focused images convert better.
2. Study leading brands for photography direction -- Dorian reverse-engineered Zara's low-angle, model-interaction photography style and applied it to the Amazon listing with a 45% click share improvement.
3. Consumer video feedback reveals what data cannot -- Ten 4-minute shopper videos uncovered issues like inconsistent pink shading across images that no one on the team had noticed.
4. Pricing inconsistency kills trust -- Having one color variation priced higher than others confused shoppers and reduced confidence in the listing.
5. Poll before you launch in fashion -- For design-led categories, spending $100 on pre-launch testing can steer you away from a bad product or nail the positioning first time.
6. Sell the transformation, not the features -- Women's fashion is an emotional, aspirational purchase. The listing should make the shopper feel "this could be me" rather than listing fabric specs.
7. Beware of over-optimizing click-through rate -- Pushing CTR too hard can lead to image fatigue and diminishing returns. Balance scroll-stopping visuals with long-term brand consistency.

Broad Match - Danny and Adam break down Amazon's financial trajectory ahead of the Q4 2025 earnings call, exploring why Prime has effectively tapped out, where the retail business is heading, and why Rufus may be Amazon's most important bet for the future of e-commerce.

Host: Danny McMillan Co-Host: Adam "Heist" Runquist

Episode Summary

With Amazon's Q4 2025 earnings call on the horizon, Adam digs into the historical financials of Amazon's retail business to understand where the company has been and where it is heading. The picture is clear: Prime membership has reached over 200 million Americans, covering roughly 75% of the adult population, and growth has slowed to just 3-4% annually. The remaining unsubscribed population is largely economically unfeasible to convert.

The numbers tell a compelling story across Amazon's retail business units. First-party retail has matured and is effectively flat or declining. Third-party seller fees have grown 190% since 2019, far outpacing the 75% growth in Amazon's own retail — but sellers are now squeezed to single-digit net margins with little room for further extraction. Advertising remains the standout at 56 billion dollars in 2024 with 300% growth over five years, yet its long-term sustainability depends on healthy seller participation.

This sets up what Adam describes as Amazon's innovators dilemma. Danny and Adam agree that Rufus represents Amazon's play to shift from a purchase destination to a product discovery and research platform, effectively competing with Google, YouTube, and Reddit for the consideration phase. The episode closes with a rallying call for sellers to focus on extreme efficiency, leveraging AI tools to optimise listings at a level of sophistication that was impossible even a year ago, and to prepare for a market where fewer sellers will survive but those who do will be significantly rewarded.

Key Takeaways

• Amazon Prime has effectively saturated the US market at over 200 million members, with the remaining population largely economically unfeasible to convert, signalling the end of Amazon's biggest historical growth engine.
• Third-party seller fees have grown 190% since 2019 compared to 75% growth in Amazon's own retail, but sellers operating on single-digit margins means Amazon has limited room to extract further on a per-unit basis.
• Amazon's advertising business pulled in 56 billion dollars in 2024 with 300% five-year growth, but its future depends on whether enough healthy sellers remain to sustain ad spend.
• Rufus is positioned as Amazon's answer to the innovators dilemma — shifting from a purchase-only platform to a product discovery and research destination to drive more visits, higher conversion, and larger basket sizes.
• AI tools now allow sellers to accomplish listing optimisation work in hours that previously took weeks, making sophisticated conversion optimisation accessible to small teams without additional headcount.
• The market is entering a consolidation phase where fewer sellers will survive, but those who maintain cash reserves, optimise ruthlessly, and adapt to the changing landscape will benefit as competitors exit.

Chapter Markers

00:00 - Introduction 00:40 - Why Amazon earnings matter for sellers 03:30 - Prime membership growth and saturation 06:22 - First-party retail maturity and decline 09:30 - Third-party seller fees hitting the ceiling 11:10 - Advertising as Amazon's growth engine 13:28 - Rufus and the discovery play 15:47 - The debate around Rufus and objectivity 19:07 - AI efficiency and listing optimisation 22:16 - Beyond keywords and single-dimension thinking 33:24 - Market consolidation and survival strategy 37:19 - Practical steps for sellers right now

Resources

• Seller Sessions Website
• Seller Sessions YouTube
• Adam "Heist" Runquist on LinkedIn
• Adam Heist YouTube Channel

```

Danny and Ritu dive deep into Anthropic's new Cowork feature in Claude Desktop - the good, the bad, and the ugly. Ritu shares a cautionary tale about file deletion gone wrong, while Danny demonstrates his custom skills pack that protects users from common pitfalls.

What You'll Learn:

• What Cowork is and how it differs from Claude Code and Claude Desktop
• Why the rm -rf command can permanently delete your files (and how to prevent it)
• How to set up deny lists in your Claude settings to protect critical files
• The power of skills and bootstrap files for consistent, reliable outputs
• Decision panels: letting Claude guide you through complex choices
• Cascade skills: research to article to slides in one automated flow

Key Timestamps:

• 00:00 - Introduction to Cowork
• 02:13 - Ritu's file deletion disaster story
• 08:37 - Understanding rm vs rm -rf commands
• 10:01 - Setting up deny lists for protection
• 16:00 - Evolution from Claude Desktop to Claude Code to Cowork
• 25:55 - Skills deep dive: orchestrator, quality gate, flow state
• 39:48 - Research cascade skill demonstration
• 43:18 - Decision panels walkthrough
• 48:58 - 2026 predictions: Ambient AI and pixel-free interfaces

Resources Mentioned:

• Danny's Skills Pack (available to listeners)
• Typora - Markdown editor ($14 lifetime)
• Time Machine backup for Mac users
• Git for version control

Connect with Ritu Java: LinkedIn

Connect with Danny McMillan: LinkedIn | Seller Sessions