Most DTC brands aren't losing customers because they stopped caring. They're losing them because they started taking them for granted.

Agnes Seville is the Head of Paid Media at Little Sleepies — a fast-growing DTC children's pajama brand with over 500,000 members in their community. In this episode of The Human Side of MarTech, Agnes breaks down the real drivers of DTC customer retention, why brand and performance marketing fail when siloed, how Meta's algorithm is forcing a creative strategy rethink, and what brands should actually measure to know if their paid media is working.

Whether you're running a lean DTC team or trying to scale paid social without burning budget, this conversation is packed with grounded, practical insight from someone doing it at a high level.

---------

WHAT WE COVER IN THIS EPISODE:

---------

→ How Agnes went from agency side to brand side — and what she learned making that switch

→ Why Little Sleepies built a 500K Facebook VIP community (and what "real" community actually means)

→ The customer loyalty mindset shift every DTC brand needs to make

→ Why brand and performance marketing teams fail when they compete instead of collaborate

→ How Meta's algorithm is penalizing creative repetition — and how to adapt

→ The attribution tools Agnes uses (Northbeam + platform data) and why nCAC is the metric that matters most

→ How to evaluate MarTech without getting burned by the implementation cost

→ What DTC brands consistently get wrong about vendors

→ The one piece of advice she'd give her younger self: "Be brave and do it scared."

---------

TIMESTAMPS:

---------

0:00 — Introduction

1:16 — Agnes's career path: from agency to brand side

4:48 — Little Sleepies and what they actually sell

6:21 — How Little Sleepies builds a 500K-member community

8:29 — What most brands misunderstand about their customers

10:10 — The over-messaging problem (the Kuru shoes example)

13:28 — How brands earn trust vs. assume loyalty

14:55 — Performance vs. brand marketing: where does each fall short?

16:52 — Why silos between brand and performance teams hurt results

19:36 — How Agnes measures holistic media performance (Northbeam + platform data)

20:53 — Is revenue the right north star metric?

22:00 — Why new customer acquisition cost is the real metric for Little Sleepies

24:25 — If forced to choose: brand or performance?

26:24 — Meta creative fatigue and the algorithm problem

29:42 — What a great e-commerce experience actually looks like

32:16 — How to evaluate marketing technology on a lean team

34:10 — The hidden cost of new tools: human capital, not just subscription fees

36:24 — Data vs. intuition in marketing decision-making

38:28 — Why human judgment still beats AI in marketing

40:44 — One piece of advice for growing a DTC brand today

43:16 — What vendors consistently get wrong when selling to DTC brands

44:17 — Advice Agnes would give her younger self

45:34 — Closing thoughts

---------

ABOUT THE SHOW:

---------

The Human Side of MarTech shines a light on the people behind the technology — the brand and agency marketers navigating complex stacks, tight budgets, and constant change. Hosted by Sean Simon, co-founder of Blurbs (trustblurbs.com) and Cogent Collective.

New episodes drop regularly. Subscribe so you don't miss one.

---------

CONNECT:

---------

🎙 Listen on Spotify, Apple Podcasts, and all major platforms:

🌐 Read the full episode post: https://themartechmatrix.com/2026/07/04/dtc-brand-customer-retention-strategies/

📋 Discover MarTech vendors without the sales pitch: https://trustblurbs.com

---------

GUEST:

---------

Agnes Seville — Head of Paid Media, Little Sleepies

---------

TAGS

---------

DTC marketing, paid media strategy, ecommerce customer retention, brand vs performance marketing, Little Sleepies, DTC brand growth, Meta ads strategy, creative fatigue Meta, Northbeam attribution, ecommerce marketing, marketing attribution tools, customer loyalty DTC, community building DTC brands, MarTech evaluation

Most marketers use Reddit personally and have no idea how to think about it professionally. This episode changes that.

Jonny Waite is the co-founder of RECHO — one of only two full-service White Hat Reddit agencies in the US. With 17 years in digital and deep roots in SEO, paid search, and omnichannel strategy, Jonny has built Reddit programs for some of the biggest brands in the world.

In this conversation, we break it down in three moves: how people actually behave on Reddit, how that translates into brand strategy, and why it's quietly shaping how AI tools like ChatGPT and Google surface your brand to the world.

What you'll learn:

→ Why 62% of people fact-check AI answers on Reddit

→ The difference between paid and organic Reddit strategy

→ How to engage in subreddits without getting banned

→ Why brands that start now will be nearly impossible to catch by 2027

→ The step-by-step process for getting started this week — not this quarter

Whether you're a brand marketer, agency leader, or just trying to figure out where Reddit fits in your media mix, this one is worth your time.

1. REDDIT IS MAINSTREAM NOW — AND MOST BRANDS ARE BEHIND

Reddit has been around for nearly 20 years but exploded in the last few years thanks to data licensing deals with Google and OpenAI. It's no longer a niche platform for tech enthusiasts. Half a billion weekly active users. 42% Gen Z. 40% Millennials. Most brands haven't figured it out yet — which is the opportunity.

2. PEOPLE GO TO REDDIT TO VALIDATE AI ANSWERS

62% of people who try to fact-check AI responses go to Reddit. That stat alone should change how brands think about the platform. It's not just a social channel — it's where trust is built or broken in the AI era.

3. AI SLOP IS DRIVING PEOPLE BACK TO HUMAN CONVERSATION

The flood of AI-generated content has made people distrust polished brand content. Reddit's anonymity and self-governance make it feel real. Brands that show up authentically benefit from that halo.

4. PAID AND ORGANIC ARE TWO COMPLETELY DIFFERENT STRATEGIES

Reddit ads are still cheap (spend is up 3x YoY but only 2% of overall media mix). Great test channel. Organic is where the real long-term value lives — but it's nuanced, time-intensive, and easy to get wrong.

5. MOST BRANDS GET BANNED BECAUSE THEY SKIP THE RELATIONSHIP BUILDING

The #1 mistake: jumping into a subreddit only when your brand is mentioned. Reddit communities see right through it. You need to be a consistent, helpful presence before you ever mention your product.

6. REACH OUT TO MODERATORS BEFORE YOU POST

Moderators govern each subreddit independently. Being transparent about who you are and what you want to contribute goes a long way. Many will let brands engage if approached respectfully and honestly.

7. START SMALL — 3 TO 5 SUBREDDITS MAXIMUM

Pick smaller, less mission-critical subreddits first. If you make a mistake and get banned, it won't tank your strategy. Use it as a learning environment before going after the big communities.

8. SOCIAL LISTENING IS YOUR FIRST MOVE

Before posting anything, onboard Reddit into your social listening platform. Understand how your brand, your competitors, and your category are being discussed. That data tells you where to show up and what to say.

9. BUILD AUTHENTICALLY — THE BENEFITS FOLLOW

Don't try to game Reddit like Google SEO. Brands that focus on genuine helpfulness — answering questions, sharing expertise, engaging in adjacent conversations — naturally accumulate AI citations, Google visibility, and qualified referral traffic.

10. THE WINDOW IS CLOSING

Jonny's prediction: brands that invest seriously in Reddit now will have an 18-month head start on competitors by end of 2027. That head start compounds — both in Reddit reputation and in AI training data. It will be very hard to catch up.

Most ecommerce brands think their retargeting is handled. They're running Criteo or Performance Max, hitting their ROAS targets, and moving on.

Here's the problem: their platform is only reaching 15–20% of the visitors who bounced from their site. The other 80–85% never see their brand again.

In this episode of Inside the Blurb, I sit down with Jason Blom — VP of Strategic Opportunities at RTB House and former Criteo sales leader — to break down what deep learning retargeting actually does differently, and whether adding a second retargeting partner is worth it.

We cover:

✅ Why no single retargeting platform can reach more than 15–20% of your eligible audience

✅ Deep learning vs. machine learning — what the difference actually means in practice

✅ Why 60%+ of RTB House-driven purchases happen on products shoppers never viewed or carted

✅ The exact framework for setting up a clean incrementality test

✅ What RTB House collects that standard retargeters don't (beyond the 5 standard pixels)

✅ How RTB House's creative philosophy differs from the ugly retargeting ads you're used to

✅ The question most brands forget to ask when evaluating a lower-funnel partner

This isn't a pitch — it's a candid conversation between two former Criteo colleagues who know this space inside and out.

Read the full breakdown + key takeaways: [INSERT POST URL]

Check out RTB House on Blurbs: https://www.trustblurbs.com

---

TIMESTAMPS

00:00 — Introduction: How Jason and Sean know each other (Criteo origins)

01:14 — Why Jason spent 7 years evaluating RTB House before joining

04:11 — The moment Jason knew RTB House technology was real

05:46 — What it's like selling against your former employer

08:45 — The Shopkick detour and what retail roots taught Jason about retargeting

13:17 — Deep learning vs. machine learning — what it actually means

16:08 — Why standard DSPs can't replicate deep learning performance

18:11 — How RTB House bids differently than Criteo and Google

22:16 — Where AI is taking performance marketing next

25:41 — The 60% stat: why most purchases come from never-viewed products

30:10 — Consumer behavior online vs. in-store (and why it matters for retargeting)

31:45 — Making the case for running two retargeting partners

35:28 — How to set up a clean incrementality test

39:15 — What it takes to get started with RTB House

41:17 — The first-party data advantage: beyond the 5 standard pixels

43:29 — How RTB House clients eventually consolidate their spend

46:21 — RTB House's creative philosophy and why retargeting ads can look better

48:52 — Rapid fire: most common objection, best question to ask, advice to younger self

52:33 — Closing thoughts

---

ABOUT INSIDE THE BLURB

Inside the Blurb is a podcast by The MarTech Matrix where Sean Simon interviews founders and leaders from the vendor side of marketing technology. No buzzwords. No fluff. Just honest conversations about what these platforms actually do and how they do it.

Website: https://www.themartechmatrix.com

Blurbs: https://www.trustblurbs.com

---

## TAGS (paste into YouTube Tags field)

retargeting, deep learning retargeting, RTB House, performance marketing, ecommerce retargeting, retargeting ads, incrementality testing, Criteo alternative, retargeting strategy, performance display advertising, machine learning vs deep learning, ecommerce marketing, DSP advertising, programmatic advertising, retargeting ROAS, martech podcast, inside the blurb, the martech matrix, Sean Simon, Jason Blom, product recommendation ads, DPA ads, ecommerce growth

Your marketing team is investing in technology — but the ROI isn't showing up. Tools go underused. Stacks overlap. Every new rollout feels like dragging people through mud.

That's not a vendor problem. It's a structure problem.

Jean Guan spent 15 years at CDW building the company's MarTech enablement team from the ground up — from a server sitting under someone's desk to a full governance function connecting marketing strategy with enterprise IT infrastructure at scale.

In this episode of The Human Side of MarTech, Jean shares:

✅ How CDW built its MarTech team from zero (and why most large enterprises get this wrong)

✅ The enterprise vendor selection process — IT, legal, procurement, finance, AND security in the same room

✅ Why most MarTech implementations fail (hint: it's not the technology)

✅ How CDW used a Center of Excellence to drive AI adoption across the organization

✅ The 3 pillars every MarTech professional needs to master: Technology, Process, and People

✅ What agentic AI means for the customer experience — and the governance challenges ahead

✅ Rapid fire: the most overused MarTech buzzword, the biggest selection mistake, and the trend to watch

If you're managing a marketing tech stack, evaluating vendors, or leading your team through AI transformation — this episode is required listening.

─────────────────────────────

TIMESTAMPS

─────────────────────────────

00:00 — Introduction: Jean Guan's background and MarTech journey

01:00 — University of Wisconsin-Madison and getting into consulting

03:00 — Joining CDW: the server under the desk story

04:00 — Why enterprises need a dedicated MarTech function

06:00 — How enterprise MarTech vendor selection actually works

08:30 — The role of procurement, legal, and security in tool selection

10:10 — Governance and driving tool adoption inside large organizations

13:00 — Change management: why it's emotional, not just a project plan

17:00 — How CDW thinks about AI and the customer experience

20:00 — AI adoption internally: CDW's Center of Excellence approach

23:00 — MarTech stack overlap and the CDP problem

27:00 — Skills every MarTech professional needs to build

31:00 — Career advice: which path into MarTech is right for you?

35:00 — Rapid fire: buzzwords, mistakes, and trends

38:00 — Why AI agents and healthcare are Jean's ones to watch

39:00 — Closing thoughts

NOTE: Adjust timestamps to match actual video timing before publishing.

─────────────────────────────

ABOUT THE SHOW

─────────────────────────────

The Human Side of MarTech is part of The MarTech Matrix podcast network — hosted by Sean Simon, founder of Cogent and Blurbs (trustblurbs.com).

Each episode goes behind the tools to shine a light on the people who use them — marketers, marketing ops leaders, and technology practitioners inside brands and agencies sharing what actually works, what doesn't, and what the industry needs to know.

🔔 Subscribe for weekly episodes

👍 Like this video if it was useful

💬 Drop your biggest MarTech challenge in the comments

─────────────────────────────

LINKS

─────────────────────────────

• Full episode post + show notes: https://themartechmatrix.com/2026/07/16/martech-vendor-selection-tips/

• Blurbs — discover marketing technology without the sales noise: https://trustblurbs.com

• The MarTech Matrix: https://themartechmatrix.com

• Connect with Sean Simon on LinkedIn: https://www.linkedin.com/in/seansimon

Most companies are either racing to bolt AI onto everything or frozen because they don't know where to start. Angela Hill does neither.

Angela runs e-commerce for Batteries Plus and has spent 20+ years in the industry — enough hype cycles to be skeptical of the next one. Her rule: test it, use it if it solves a real problem, fire it if it doesn't.

In this episode of The Human Side of MarTech, Angela breaks down:

✅ Why AI doesn't create new problems — it exposes the ones you already had (messy data, disorganized content, broken search)

✅ Her team's 90/10 rule for using AI as an assistant, not a replacement

✅ How she filters vendor pitches using the "fitment" test in a market of 20,000+ MarTech vendors

✅ Why franchise e-commerce (BOPIS, franchise owners, urgent-need customers) breaks the standard MarTech playbook

✅ What separates a vendor pitch that earns her attention from one that wastes her time

✅ What she actually looks for in an "AI-native" hire

This isn't hype or fear — it's a grounded look at how an experienced operator adopts technology on her own terms.

📖 Read the full breakdown + key takeaways: [INSERT POST URL]

🔗 Compare MarTech vendors on Blurbs: https://www.trustblurbs.com

─────────────────────────────

⏱️ TIMESTAMPS

─────────────────────────────

00:00 — Introduction: Angela's 20+ years in e-commerce

02:00 — The biggest surprise across every hype cycle: the fundamentals never change

05:49 — Walking into Batteries Plus: legacy tech debt and where to start

09:07 — How she audited the existing tech stack

10:56 — AI adoption: from resistance to "we have to test this"

13:15 — What AI is really exposing: messy data, disorganized content

14:22 — Agentic search and preparing the business to be found

14:43 — How AI is used day-to-day across the team

18:13 — Why franchise e-commerce is a different game

20:36 — The "fitment" test: how she filters vendor pitches

24:09 — Why standard personalization playbooks don't work for BOPIS

25:52 — From 250 vendors to 20,000: how to even know where to start

29:10 — Why she relies on conferences and peer networks over cold research

32:57 — How a lean team stays aligned on what tools everyone's using

34:20 — The difference between a pitch that respects her time and one that wastes it

38:17 — What trade shows get right — and where they could improve

42:24 — What she'd want every AI vendor to know before pitching her

44:40 — What an "AI-native" hire actually looks like

47:29 — Closing thoughts

─────────────────────────────

🎙️ ABOUT THE SHOW

─────────────────────────────

The Human Side of MarTech is part of The MarTech Matrix podcast network — hosted by Sean Simon, founder of Cogent and Blurbs (trustblurbs.com).

Each episode goes behind the tools to shine a light on the people who use them — marketers, marketing ops leaders, and technology practitioners inside brands and agencies sharing what actually works, what doesn't, and what the industry needs to know.

🔔 Subscribe for weekly episodes

👍 Like this video if it was useful

💬 Drop your biggest AI adoption question in the comments

─────────────────────────────

🔗 LINKS

─────────────────────────────

• Full episode post + show notes: [INSERT BLOG POST URL]

• Blurbs — discover marketing technology without the sales noise: https://trustblurbs.com

• The MarTech Matrix: https://themartechmatrix.com

• Connect with Angela Hill on LinkedIn: [INSERT ANGELA'S LINKEDIN URL]

• Connect with Sean Simon on LinkedIn: https://www.linkedin.com/in/seansimon

Raf Lopez de Azua graduated from the Naval Academy months before 9/11. By the time he got to his ship, everything had changed. What followed was a career built across government consulting, a Cornell MBA, global brand training at P&G, and now Coty — where he runs global media effectiveness and serves as the company's AI champion.

In this episode, we get into:

→ What 22-year-old wartime leadership actually teaches you about managing teams

→ Why "nobody's gonna die" unlocks better decision-making in marketing

→ How P&G's training is worth more than an Ivy League MBA

→ The real story behind Coty's GMP consolidation — and why it didn't work

→ What vendors get wrong about enterprise tech debt and switching costs

→ How Coty governs AI with a cross-functional board and a buy/rent/build framework

→ Why agentic buying is intriguing — and the question nobody's answering yet

→ The agency talent pipeline problem nobody wants to talk about

If you buy media at scale, manage a global marketing org, or sell into enterprise accounts, this one is essential.

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

CHAPTERS

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

00:00 From Naval Academy to 9/11: The Origin Story

02:12 What Military Leadership Actually Teaches You

05:00 From Navy to Brand Marketing — The P&G Connection

10:58 Global Strategy vs. Local Reality at P&G

17:54 Resolving Global-Local Tension With Data

20:57 The GMP Consolidation Promise (And Why It Failed)

25:39 Learning to Live With Measurement Uncertainty

29:07 What Vendors Get Wrong About Tech Debt

36:05 Shifting From Efficiency to Effectiveness

42:42 Inside Coty's AI Champion Role

45:10 AI Governance: The Buy/Rent/Build Framework

48:09 Agentic Buying and the Future of Media Agencies

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

KEY TAKEAWAYS

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

1. Military leadership = managing whole people, not just tasks

2. "Nobody's gonna die" — perspective is a performance multiplier

3. P&G brand training outperforms any MBA for practical marketing skills

4. Global brand consistency: non-negotiable pillars + data-driven local flexibility

5. The unified tech stack is a myth. Walled gardens won't consolidate.

6. Perfect measurement data doesn't exist — learn to work with uncertainty

7. Vendors: enterprise switching costs are real. Get into the consideration set early.

8. Build efficiency first. Then layer effectiveness on top.

9. AI governance belongs to IT, legal, data, AND marketing — not just one team

10. Agentic buying: intriguing concept, but who catches the hallucinations?

11. Thinning the agency talent pyramid with AI may break the expert pipeline

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

CONNECT

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

🌐 Full episode + transcript: TheMarTechMatrix.com

🔗 Discover MarTech vendors: trustblurbs.com

📧 Work with Cogent: cogentcollective.com

🎙 Subscribe wherever you listen to podcasts

#MarTech #GlobalMarketing #AIChampion #TheHumanSideOfMarTech #MartechMatrix #ProgrammaticAdvertising #MediaEffectiveness #AgenticAI #Coty

Last-click says one thing. Your CFO doesn't believe it. Your incrementality test says something else entirely.

None of those numbers are wrong — they're just not measuring the same thing.

In this episode of Inside the Blurb, I sit down with Madan Bharadwaj, founder and CEO of M-Squared. Madan has built four companies inside the attribution and measurement space, including Visual IQ (one of the first multi-touch attribution platforms) and Measured (the industry's first pure-play incrementality company).

We cover:

-Why nearly everything brands use to measure marketing is "correlational" and carries built-in bias .

-Why incrementality testing is deceptively hard to run well

-Triangulation — turning conflicting attribution, MMM, and incrementality numbers into one multiplier and one decision

-A real story about an AI model recommending a brand cut $25M in Amazon ad spend, and why it was wrong

-Why causal data has to be built, not mined

-Who actually needs this level of measurement (and who doesn't)

Read the full breakdown + key takeaways: www.themartechatrix.com

Check out M-Squared on Blurbs: https://www.trustblurbs.com

TIMESTAMPS

00:00 — Introduction: welcome to the Matrix

01:00 — Madan's origin story: from computer science to a $10M/month PPC bidder

07:01 — 2005 vs. 2026: what's changed in measurement, and what hasn't

10:55 — How GDPR, CCPA, and walled gardens shrank trackable behavior

13:07 — Why brands still lean on last-click even though it's incomplete

16:10 — Why incrementality testing sounds simple and is actually hard

18:30 — Designing a "feasible test" before you run it

20:07 — Do marketers use data to confirm what they already believe?

24:42 — Why the best marketers combine data with business instinct

26:45 — Causal vs. correlational, defined

30:08 — What makes M-Squared different

33:51 — Triangulation, explained

38:38 — The causal insights program

41:38 — Who actually needs advanced measurement

42:58 — Why AI can't manufacture causal data

44:07 — The $25M Amazon ad spend story

46:56 — Why domain expertise still matters

47:10 — Squared Collective: next workshop September 10, Boston

49:10 — Closing thoughts

ABOUT INSIDE THE BLURB Inside the Blurb is a podcast by The MarTech Matrix where Sean Simon interviews founders and leaders from the vendor side of marketing technology.

Website: https://www.themartechmatrix.com

Blurbs: https://www.trustblurbs.com

LinkedIn: https://www.linkedin.com/in/madanbharadwaj/

Most vendor outreach fails before it's even opened. And most demos lose the deal before the second slide.

In this episode of The Human Side of MarTech, Sean Simon sits down with Chris Seahorn, Co-founder of Boundless Digital — a fractional CMO agency he started two years ago with a longtime colleague after 25+ years in digital marketing at companies like eBags, MapQuest, Naked Wines, and others.

Chris evaluates MarTech stacks across multiple clients simultaneously. He has an unusual vantage point: he sees what vendors do well, what kills deals, and what separates implementations that stick from ones that quietly die at renewal time.

In this conversation:

→ The StoryBrand framework: why "your customer is the hero" changes everything about how you sell

→ What makes Chris immediately delete a cold email (and what gets a meeting)

→ The "kitchen sink demo" problem — and how to fix it before you walk in the room

→ How a fractional CMO evaluates a client's tech stack from day one

→ Why conversion rate optimization is always the first quick win (and how 5% improvements stack to 25–30% total lift)

→ The most common checkout conversion killers: coupon confusion, account creation walls, and traffic mismatches

→ Why SaaS implementations fail — and it's almost never the product

→ AI customer service agents: why they went from "worse than useless" to genuinely impressive

→ The self-checkout analogy and what it tells us about AI adoption timelines

→ What Chris would tell his 1999 self: "Progress over perfection — trust your gut, but come with the receipts"

0:00 — Introduction: Chris Seahorn and Boundless Digital

1:30 — Chris's 25-year journey: eBags, MapQuest, Naked Wines, startups

3:30 — Why Chris left the CMO seat to go fractional

5:30 — What signals to Chris that a vendor actually understands his world

8:00 — The StoryBrand framework: hero, guide, problem, plan

11:00 — The "creep factor" — when vendor research goes too far

13:30 — Getting from inbox to meeting: breaking through the clutter

16:00 — What makes a good vs. bad vendor demo

19:30 — The kitchen sink problem: analysis paralysis in presentations

22:00 — Why use cases and follow-ups close more than the deck does

24:30 — How a fractional CMO walks into a new client

27:00 — The first assessment: looking at data before charging anything

30:00 — Evaluating a tech stack: measurement first, then fundamentals

33:00 — Email/SMS, CRO, and A/B testing as the foundation

36:00 — Conversion rate: the tide that lifts all ships

39:30 — Why 5% improvements on key pages beat swinging for 50%

42:00 — The most common conversion killers (coupon confusion, account walls)

45:30 — Seasonal behavior: why November cart abandonment is actually a good sign

48:00 — The tire store checkout story — friction that kills conversions

51:00 — Are consumers actually changing their behavior with AI?

54:00 — AI recommendation engines and what winning e-commerce brands do

57:00 — AI customer service agents: from worse-than-useless to genuinely impressive

60:00 — The self-checkout analogy and AI adoption timelines

63:00 — Why SaaS implementations fail: no champion, no success definition

66:00 — Chris's advice to vendors: step out of your own shoes

68:00 — The case for trials and reduced-rate pilots

70:00 — Advice to his early-career self: progress over perfection