Human in the Loop

This brand new show explores the dynamic intersection of human intelligence and cutting-edge technology within the ever-evolving landscape of ecommerce.From in-depth discussions on artificial intelligence, to viral brands that launched on TikTok, to the stories behind today’s biggest technology companies, brands, and sellers..Welcome to the forefront of the ecommerce revolution, where being a Human in the Loop is not just a just a strategy...But the very essence of progress.

Human in the Loop is a business podcast hosted by Teikametrics, with 54 episodes on record and a Required Pod Score of 80. PitchCentric scores this show on Booking Probability, Listen Score, and live audience signals refreshed every 24 hours.

Teikametrics hosts Human in the Loop, a business show with 54 episodes published.

Our AI reads these to draft pitches. Use them as grounding for a pitch that cites a real guest and a specific topic.

• 

  Human in the Loop

  Agentic Commerce is Here & Most Brands Aren't Ready

  Aug 19, 20266 min

  Amazon just rewrote the rules of the product listing — and it happened almost overnight. In this eTail Boston conversation, Cameron Yoder sits down with Teikametrics founder and CEO Alasdair McLean-Foreman to unpack Amazon's July 27th mandate capping every product title at 75 characters and introducing a brand-new "item highlights" field. The real story isn't the character count — it's what it signals: Amazon is racing to power Alexa for shopping, where Andy Jassy revealed on the Q2 earnings call that consumers are already buying 40% more. Alasdair breaks down how brands can prepare for agentic commerce across Alexa, Walmart's Sparky, ChatGPT, and Gemini, why structured content is now table stakes, and how TikTok is pushing the boundaries of AI-generated content. If you sell online, this is the wake-up call on AI discovery you didn't know you needed.Key TakeawaysThe 75-character title change is a signal, not a formatting tweak. On July 27th Amazon capped every product title at 75 characters and added an "item highlights" field. Alasdair's point: this isn't cosmetic — it's Amazon restructuring listings to feed Alexa for shopping and agentic buyers.Agentic commerce is already moving volume. Andy Jassy said on Amazon's Q2 earnings call that consumers using Alexa for shopping are buying 40% more than other customers. That's not a future projection — it's happening now, and Q4 could see more agent-driven transactions than ever.Alexa for shopping is the North Star for the next 6–12 months. With Prime, mobile-app integration, and the sheer volume Amazon commands, Alexa is putting up the biggest numbers among AI discovery platforms — and mobile searches are already converting into Alexa-for-shopping searches.Structured content is now table stakes. To be found by an AI agent, you have to feed it structured content. Ads remain a closed-loop, pay-to-play channel that isn't going away — but brands that don't invest in content to feed the agent will get left behind.The AI discovery race has several horses. Alexa, Walmart's Sparky, ChatGPT, and Gemini are all in play. The open question every brand operator should be asking: which platform becomes the primary place shoppers discover new products?TikTok optimizes for content, not the product detail page. TikTok's growth is faster than any platform they've seen, and TikTok Shop is where content is king. The optimization lever isn't the PDP — it's the creator content itself, and increasingly the AI used to make and remix it.AI is moving from text to video and images. Beyond titles, bullets, and copy, tools like TikTok's Symphony Gen AI suite are generating better creator content — and even fully AI-generated content — while predicting what will resonate with a given audience.Chapters0:00 Introduction — Enterprise Brands at eTail Boston0:28 The 40% Stat: Alexa for Shopping Takes Off2:24 Inside Alasdair's Keynote2:41 Amazon's July 27th Title Change & Item Highlights2:56 Why This Is Really About Agentic Commerce4:33 How Brands Get Ready: Content Is the New Frontier5:00 The AI Discovery Race: Alexa, Sparky, ChatGPT, Gemini5:45 Where TikTok Fits In6:35 AI-Optimized Video, Images & TikTok Shop

  ListenShow Notes
• 

  Human in the Loop

  Teikametrics MCP Explained: Connecting Amazon, Walmart & TikTok Data to Claude

  Jul 15, 202616 min

  What if your e-commerce data could talk directly to Claude? In this episode, Teikametrics CEO & Founder Alasdair McLean-Foreman breaks down the launch of the Teikametrics MCP — and why the team deliberately built it around proprietary optimization, content, and catalog data rather than simply mirroring what already lives in Amazon, Walmart, or TikTok. Alasdair explains where LLMs genuinely add value (stitching together siloed data from inventory, finance, and content systems) and where they don't (directly optimizing your ad levers — that's what a vertical AI orchestration layer is for). The beta launches at the end of July — listen in to learn how early users are already putting it to work, and how to become one of them. Key Takeaways Proprietary data is the point. Amazon's MCP already exposes Amazon's data. Teikametrics built its MCP around what only lives in its platform — optimization data, content recommendations, and catalog outputs — so customers unlock signals they can't get anywhere else. LLMs are for stitching, not steering. Alasdair is clear that customers shouldn't use Claude to directly optimize Amazon, Walmart, or TikTok ads — that requires a vertical AI orchestration layer, which is what the Teikametrics platform does. The MCP shines when combining Teikametrics data with inventory, finance, or content systems. The Ask Elephant example set the template. Teikametrics saw the power of MCP internally by piping sales-call transcriptions from Ask Elephant into Claude and stitching them with product data — and set out to replicate that experience for customers. Catalog review at scale is a killer use case. With Amazon's July 27 content title change requirement, brands can pull Teikametrics' AI catalog recommendations into Claude and check them against their own brand voice and IP guidelines — practical when reviewing 10,000 SKUs in a UI isn't. Customers will find the use cases. The beta is a deliberate "wide door": Teikametrics expects design partners to discover applications the team hasn't imagined, and those discoveries will feed back into the core product roadmap. "We have AI" is a dead marketing message. At Cannes, generic AI banners have given way to companies showing how AI is actually being used — and MCP is a concrete example of functionality over hype. Read first, write carefully. V1 of the MCP surfaces data out of the platform. Read-write capabilities — letting Claude make changes inside Teikametrics — are on the roadmap, but will be rolled out cautiously with design partner feedback. 0:00 — Introduction: Exposing proprietary data, not marketplace data 0:49 — Dispatch from Cannes: Amazon's takeover, Walmart's CTV push, TikTok's creator economy 3:03 — Why "we have AI" doesn't land anymore 3:55 — Introducing the Teikametrics MCP 4:51 — The Ask Elephant + Claude example that inspired the strategy 6:15 — Use case: reviewing 10,000 SKUs of catalog content in Claude 7:49 — What the MCP does (and doesn't) do in v1 9:20 — Why not just use Amazon's MCP? The orchestration layer argument 11:33 — Beta launch, design partners, and what success looks like 13:09 — Breaking down data silos: inventory, finance, and reporting 14:40 — Timeline: end-of-July release and the read-write roadmap

  ListenShow Notes
• 

  Human in the Loop

  Will AI Replace Creators? Inside TikTok's Symphony Suite & Seedance 2.0

  Jul 1, 202621 min

  The internet is about to be flooded with AI-generated content — so what actually breaks through? In this episode, Cam sits down with Ali Tatarzyn, Director of Product at Teikametrics, to unpack TikTok's AI creative suite: Symphony, the advertiser toolbox, and Seeance 2.0, ByteDance's own multimodal video model now powering it. They dig into whether AI will replace creators (short answer: it replaces low-differentiation content, not earned trust), why TikTok's real constraint is creative supply rather than ad inventory, and how the game shifts when content becomes nearly free to produce. Ali also shares a prototype TikTok creative brief generator he built with Claude — and the surprising lesson that AI output is only as good as the strategy behind the prompt. Listen in for a clear-eyed view of where brands win, and where they're about to drown in mediocre content. Key Takeaways AI replaces content, not creators. Ali draws a sharp line between a creator — a person with trust, taste, distribution, and a community — and content as a disposable asset brands need to test. AI is excellent at the latter (hook variations, product demos, explainers, avatar content) but can't manufacture earned trust. The flood makes the best creators more valuable. As average-looking AI content floods every feed, the people who break through with humor, specificity, and real community become scarcer and more valuable — not less. TikTok's real constraint is creative supply, not ad inventory. Budgets and media plans don't matter if the creative is stale or doesn't feel native. TikTok rewards freshness, volume, and cultural fit, and creative fatigue hits fast — which is exactly why TikTok is building AI tools to lower the cost of producing native content. The shift is from "help me think of an ad" to "help me make 20 of them." A year ago AI gave you hooks and scripts upstream; you still had to shoot and edit. Now Symphony and Seedance generate finished, ready-to-run assets — turning one product image or existing video into 5, 10, or 20 short-form variations. Teikametrics' lane is intelligence and measurement, not video generation. Ali is explicit that competing as another video model is a losing, commoditizing battle. The opportunity is telling brands which products to push and which hooks to test (upstream), then measuring whether content drove TikTok Shop GMV, branded search lift on Amazon, or real SKU movement (downstream). Cheaper content makes measurement harder, not easier — that's where Halo comes in. When brands jump from testing 5 videos to 50, views aren't enough. TikTok Halo is the measurement layer that answers what each piece of creative actually did for the business and what to make more of. A bad strategy executed perfectly is still a bad strategy. Ali built his brief generator first — not a video tool — because the brief is where the thinking happens. Generate 100 videos from a weak brief and you've just made 100 bad videos faster. Quality jumped most when he treated the prompt like a real product strategy, not a simple command. 0:00 — Cold Open: Volume vs. Quality 0:35 — What Are TikTok Symphony and Seedance 2.0? 1:37 — Will AI Replace Creators? 3:13 — Why Trust Is the Thing AI Can't Fake 5:00 — TikTok's Real Constraint: Creative Supply 6:43 — How Symphony Lowers the Cost of Native Content 8:15 — From Creative Ideas to Finished AI Assets 10:53 — The Opportunity for Teikametrics 13:16 — What This Means for TikTok Halo 14:41 — Building a Creative Brief Generator with Claude 17:05 — Why Start With the Brief, Not the Video 19:10 — What Brands Should Watch Out For 19:53 — Closing Thoughts

  ListenShow Notes
• 

  Human in the Loop

  Is Prime Day Still Worth It? How to Approach Prime Day in 2026

  Jun 18, 202616 min

  Prime Day is bigger than ever — but is it still a must-win moment, or just another peak in an always-on promotional calendar? In this episode of Human in the Loop, Teikametrics Premium Services analyst William Fleisher breaks down how Prime Day has evolved into a multi-day, omnichannel event spanning Amazon, Walmart, and beyond. He explains why the brands that win treat Prime Day as a window rather than a date, how to use DSP and streaming TV across the lead-in and lead-out, and what success actually looks like in 2026. Whether you're a seller weighing your budget or a shopper wondering if the deals are real, this conversation cuts through the noise with practical strategy. New episode is live just in time for Prime Day. Chapters: 0:00 — Introduction — Will's content debut 0:21 — Is Prime Day still a must-win moment? 1:52 — Prime Day as an omnichannel event 3:04 — Are the deals actually the best of the year? 4:32 — What deals look like this year, brand by brand 5:46 — What does success look like for Prime Day 2026? 7:00 — The biggest mistake brands make 8:24 — How to leverage DSP for Prime Day 9:56 — Streaming TV and other ad types 11:17 — What's surprised Will most this year 13:19 — The future of Prime Day: 2027 and beyond 15:12 — Wrap-up Key Takeaways Prime Day is still a must-win — even in an always-on world. The promotional calendar is more crowded than it used to be, but with record sales last year, treating Prime Day as "just another peak" is the wrong framing. It remains a major launching pad if you commit to it. It's an omnichannel event now, not just an Amazon one. With Walmart and Target running simultaneous deals, consumers no longer assume Amazon has the best price. The brands getting the most out of the event are running a concerted strategy across retailers — not just checking a box. The deals are still real — they just look different. Headline price cuts on big-ticket electronics may be smaller than years past, but discounts have spread to everyday consumables, where a few dollars off can be a large percentage. You sometimes have to dig a little more. Success is measured over a window, not a silo. With a longer event and an omnichannel footprint, sophisticated brands track Prime Day performance across the full lead-in and lead-out period — and manage budget carefully so it lasts all four days, including the final-day spike. The biggest mistake is focusing only on the event days. Doing the work early compounds results. Sponsored Brands, Sponsored Display, and DSP in the lead-in set you up for success, and many of those same tactics pay off in the lead-out. DSP shines in the lead-in and lead-out. Use it beforehand to build awareness and stay top-of-mind, so your lower-funnel Sponsored Products convert better on the day. Afterward, retarget recent purchasers or shoppers who viewed your PDP but didn't buy — at a cheaper cost than during peak Prime Day competition. Streaming TV is intriguing — but earn it first. It's a newer tactic now open to more brands, but only worth it once your fundamentals are in place: strong creative, good inventory, solid reviews, and a page that converts. If you're already running Sponsored Products and Brands well, those boxes are likely checked. Prime Day isn't slowing down. Expect events to keep growing and expanding from big-ticket items to every product category, with content distribution across platforms like TikTok adding even more ways to drive demand.

  ListenShow Notes
• 

  Human in the Loop

  Why Is Amazon Cutting Your Product Titles in Half?

  Jun 17, 202613 min

  Amazon just dropped a bombshell on sellers: product titles are being capped at 75 characters (down from the 200–250 most brands rely on) with a new comma-separated "Item Highlights" field absorbing the extra details. In this episode, Ryan Goldstein unpacks what's actually changing, why Amazon is doing it (think mobile-first design, AI-driven discovery, and structured data that's easier for LLMs to query), and what it means for the title real estate sellers have spent years optimizing. The catch: the deadline is July 27, and if you don't make the changes yourself, Amazon will make them for you. Listen for a practical game plan on how to prepare without disrupting your listings right before Prime Day. Chapters : 0:00 — Introduction: An Amazon title bombshell 1:02 — What exactly is Amazon changing? 1:58 — The July 27 deadline (and what happens if you do nothing) 2:20 — Why Amazon is making this change: mobile, AI & structured data 5:30 — What it means for sellers and brands 7:49 — How to prepare: previewing and planning your changes 8:55 — Doing it at scale with Teikametrics Catalog & Smart Pages 10:12 — Timing it right: why Prime Day changes the playbook 12:14 — Wrap-up and what's next Key Takeaways Titles are being cut to 75 characters — but you don't lose the detail. Amazon is splitting your old title into two parts: a short, scannable title and a new "Item Highlights" field that adds roughly 125 characters for use cases and product specifics. Item Highlights must be structured, comma-separated data. Amazon is explicit that highlights can't be one long string — they have to be comma-separated, which turns your listing into structured data that's far easier for Amazon to query and surface. This is really about mobile and AI discovery. Shorter titles create a cleaner mobile interface where Amazon sees shopping trending, and the structured highlights feed AI models that increasingly power product discovery. The July 27 deadline is a hard line in the sand. Amazon won't shut off your listing if you do nothing — but it will rewrite your title and highlights for you, and the auto-generated version often gets the order and emphasis wrong. Your title is prime real estate you don't want to hand over. Title and main image are the two things that earn the click. With only 75 characters, brands face new decisions — like whether a well-known brand name even belongs in the title anymore. You can preview the changes right now. Sellers can already see Amazon's proposed title and highlights in the console before the change takes effect — a smart first step to spot where the auto-suggestions miss. Don't make big listing changes heading into Prime Day. With Prime Day next week, the advice is to build your plan now but hold major title edits until July, so you don't disrupt indexing during one of the biggest e-commerce events of the year.

  ListenShow Notes

Based on semantic analysis of episode topics and host coverage, this show is a strong guest fit for executives in:

Industry fit is computed by PitchCentric using vector embeddings of the show's episode catalog.