• LinkedIn
• Twitter
• Instagram
• YouTube
• RSS Feed

Apple •Spotify• Pocket Casts •Youtube •Overcast •RSS

What’s up everyone, today we have the pleasure of sitting down with Christina Garnett, Fractional Chief Customer and Communications Officer at Neuemotion.

Summary: Christina spent 5 years teaching math before she became a customer marketing expert, and it shows in how ruthlessly she takes apart the personalization playbook. She makes the case for core memories a brand can’t buy, walks through what actually happened when Spotify Wrapped got the AI treatment, and introduces earned context, the idea that having someone’s data and having the right to use it are 2 completely different things. Along the way she calls CX and brand the same job, refuses to bet her life on NPS, and explains why the B2B move nobody is making costs almost nothing. If you have ever shipped a perfectly segmented campaign that landed with a thud, this one will tell you why.

In this Episode…

• Why Your Tech Stack Cannot Fix Siloed Customer Teams
• Why Getting the Name Right No Longer Counts as Personalization
• Why A/B Testing Programs Ignore Qualitative Customer Data
• How Brands Build Core Memories Customers Cannot Buy
• How B2B Brands Build Core Memories Without a Budget
• Why Personalization Feels Predatory Even When You Have Consent
• Where Influence Ends and Manipulation Starts in Marketing
• Why Audiences Are Going Back to the Humanities in the AI Era

About Christina

Christina Garnett is a fractional Chief Customer and Communications Officer at Neuemotion and the founder and principal at Pocket CCO, where she builds customer trust systems for agencies and SaaS companies. She’s the author of Transforming Customer-Brand Relationships, which won the Independent Press Award and the International Impact Book Award and was a Foreword INDIES finalist, and she’s currently writing a second book built on the Customer Trust Equation she created.

Before going independent she built customer advocacy at HubSpot, growing the HubFans program from 0 to 37,000 members and more than 4,000 active advocates, and she led social and customer strategy at global scale through ICUC. She’s a contributing writer at Campaign US, was previously a byline at Adweek, and she started her career as a teacher.

Why Your Tech Stack Cannot Fix Siloed Customer Teams

Every few years a software category shows up promising to end the silo problem. The CRM was going to do it. Then the customer data platform. Then the customer 360, the composable stack, and whatever the warehouse-native vendors are calling it this quarter. Marketing, support, social, and CX all get wired into the same tables, and everyone assumes shared data will produce shared behavior. Buy the integration, get the alignment.

Christina has been on enough of those teams to know how that ends.

“A lot of brands will think that their tech stack is supposed to marry those teams together, and that’s gonna solve the problem, and it never does.”

She came into marketing sideways, an English major who taught math for 5 years, including a stretch at a school built for kids with learning differences where she trained in the Orton-Gillingham method. That method exists to move information from short-term into long-term memory. Marketers have a word for the same thing and call it brand recall. Her husband told her early on that she should go into advertising because she couldn’t watch an ad without rebuilding it in her head, tweaking the copy, cutting the music, fixing whatever was broken. She went and took every course, read every book, and kept the teacher’s instinct the whole way through: make the hard thing make sense, find the right words for the room you’re in.

Then she sat at basically every table that claims part of the customer, working as a social listening strategist, a social media creator and a social media strategist, and putting time into support, CX, community and advocacy. What she kept finding was the same organizational failure repeating under different logos, and no amount of shared tooling moved it.

Here’s the argument that runs under her whole book. Nobody is solving customer experience alone, because nobody is the only person reaching the customer. Support talks to them, so support gets blamed for the relationship. But your UX team is shaping what the customer can and can’t do, whether they think of that as CX work or not. Your legal team is writing the fine print the customer reads at the worst possible moment. Finance sets the refund window. All of it builds the relationship, and most of it never appears on anyone’s CX dashboard.

The result is a company that’s genuinely excellent in one place and quietly hostile in another, and it shows.

“You have brands that are really good at social, but their sales team is mean. Or the social team is non-existent, but their community team is killing it. It’s all over the place when you need this team sport mentality.”

That’s why the material became a book rather than a framework or a course. A framework asks you to already agree with its premise before you’ll use it. Christina wanted the thing she’d been hunting for on shelves and never found, written for the people at all of those different tables at once. Her words on it: “I think we all wind up writing the books that we wish already existed.” The structure follows the problem in 3 moves:

1. Every moving piece that actually builds a brand relationship, most of which sits outside the CX team.
2. Community as the silo killer, the thing that finally forces those teams into the same room.
3. What happens when something lands that you had no way of seeing. A brand crisis, an acquisition by a company your customers hate, or AI arriving and throwing the whole operating model in the air.

That third one is where most companies are living right now, and it’s the least rehearsed. The martech buying cycle keeps funding integration projects because integration is something you can purchase, and cross-functional accountability is something you have to negotiate. Until one person owns the customer relationship end to end on paper, a new platform just gives 4 teams a shared place to file work they still refuse to coordinate.

Key takeaway: List every team that shapes your customer’s experience, including the ones that never speak to a customer. Legal writes the cancellation language, UX decides how buried the downgrade button is, finance sets the refund window, and none of them get counted. Put all of them on one page and mark who currently owns the customer’s experience of each moment. The blanks are your real CX roadmap, and no platform purchase will fill them in for you.

Back to the top ⬆️

Speaking of tech that’s ready for an upgrade… Drift Email is being sunset on January 31, 2027. Salesloft is recommending allGood to their customers, and they’ve built tooling that can import your existing Drift Email setup while upgrading it from keyword rules to AI that understands reply intent. [Check it out] 

If you want the receipts, Robyn Hatfield is Director of RevOps at Gravyty and a former Marketo Champion. She ran Drift Email and moved to allGood. Here’s what she said, in her words:

“Anybody in marketing operations always hears from a vendor, ‘seamless integrations, plug-and-play.’ It’s like nails on a chalkboard, because it’s usually just absolute nonsense. This was the easiest migration ever. I kept thinking, am I missing something? Is this not going to work, because this is just too easy? It worked from the get-go.”

Here’s the highlight reel.

Why Getting the Name Right No Longer Counts as Personalization

Accuracy in a lifecycle program has never been cheaper. You can segment by behavior, by lifecycle stage, by predicted churn, by what someone hovered over last Tuesday, and the tooling will do all of it on a mid-tier plan. Ask a marketer whether their email is personalized and they’ll show you the merge fields and the branching logic and say yes.

Ask the person receiving it and you get a different answer. Christina’s position is that martech made accuracy so cheap that creativity started to look like a nice-to-have, and customers noticed long before marketers did.

“Getting the name right is table stakes, and so that no longer reads as personalization to people. That reads as, ‘You should know. You should have my name right.'”

Her frame for the whole industry is the Jeff Goldblum line from Jurassic Park, the one about scientists so preoccupied with whether they could that they never stopped to ask whether they should. We’re living inside that meme, she says, running on a lot of we could so we can. And the cost shows up somewhere marketers rarely look, which is in what we’ve trained people to expect.

Conditioning is real, and it doesn’t require a marketing degree. Somebody who has never opened a martech vendor’s website in their life still gets a physical wave of irritation at the 5th version of “oops, we made a mistake on our Black Friday sale, so we’re extending it.” They haven’t deconstructed a single ad campaign. They just recognize the shape of the thing. As Christina puts it, “Y’all are not serious. Do better.”

A decade ago, building a smart list so that every first name in a 40,000-person send came out correct was genuinely impressive work. Now it earns you nothing.

“Congrats on not getting fired for doing the least.”

What Chasing the Next Hook Gets Wrong About Storytelling

When the mechanical stuff stops working, most teams go looking for the next mechanical thing. Christina’s example is the entire genre of content telling you that your problem is your hook. She hates watching those videos, and her objection is that the hook is a piece of storytelling craft that predates every one of us. It’s a dark and stormy night. It’s once upon a time.

Take “put a finger down.” That opener works because it primes you to expect something a little salacious and tells you what your job is for the next 30 seconds. The priming does all the work. Reverse-engineer only the words and you’ve copied the surface of a technique whose actual engine is human behavior.

Which is why she keeps pushing marketers back toward behavioral psychology and sociology rather than toward the next format. People aren’t making decisions in a vacuum, and understanding the world those decisions happen inside beats memorizing a sequence of opening lines. Marketing has spent 10 years reinventing the wheel, she argues, when the basics were sitting right there the whole time.

When a Personalized Campaign Has No Human Behind It

The version of this that should sting for anyone who has run a lifecycle program is the fully built drip that a customer cannot escape. Everything is personalized. Every branch fires correctly. And the moment something goes wrong, the customer gets dropped into another queue with no way to reach a person.

“It’s just another campaign. It’s just another drip campaign. It’s just another workflow. It’s just another automation. It’s great when it’s paired with, ‘I can get to a real human if there’s an issue.’ But it is not a replacement for care.”

Phil has fought this exact battle in-house, arguing with IT teams about no-reply addresses on campaign sends. His most reliable ally in that fight was always someone from the CS side, because they were the ones who understood what happens on the other end of a message nobody can answer. Automation earns its place when a human is reachable behind it. On its own it just makes the wall smoother.

Then there’s the question of where the data came from in the first place. Everybody has had the experience of saying something out loud, never typing it anywhere, and getting an ad for it 2 days later. Technically that’s targeting doing its job. The first time it happens it’s interesting.

“The 10th time it happens, you’re wearing like the cutest tinfoil hat that you’ve made for yourself.”

Marketers know how the sausage gets made, so we grade this stuff on a curve. Then it happens to us and it feels exactly as invasive as it does to everyone else. The personalization arms race has been running on a metric that quietly stopped measuring anything about a decade ago. Every extra merge field buys less than the one before it, and the returns flatlined the moment customers learned to read the pattern. Whatever advantage is left in lifecycle marketing is going to come from restraint and from creative work a competitor cannot clone by copying your segment logic.

Key takeaway: Pull your last 10 lifecycle sends, strip the logo, and hand them to someone outside marketing. Ask which brand sent each one. Anything they can’t place is running on accuracy alone and is doing nothing for you that a competitor with the same tool can’t do by Friday. Fix the ones that fail by adding a point of view, a real sender, and a reply address that reaches a person.

Back to the top ⬆️

Why A/B Testing Programs Ignore Qualitative Customer Data

In most companies, quantitative and qualitative data sit in different rooms with different owners. Statistical significance and lift percentages go to one team. Customer interviews, open-text survey responses, and social sentiment go to another, if they go anywhere at all. Testing programs get built entirely out of the first pile, and the creative team gets handed the results as a verdict.

Christina’s diagnosis of that split is that it’s fear wearing a lab coat.

“If you are really data-focused and you’re not looking at the qualitative at all, it’s because you don’t want nuance.”

Nuance is where the uncomfortable stuff lives. Nuance is the customer who says “I absolutely love your brand, but here’s what I would fix.” Numbers give you a clean directional answer you can put on a slide and defend in a QBR, and they will never hand you that sentence. The person who never has to read the messy version never has to change anything.

Her example of how this plays out in practice will be familiar to anyone who has run a survey program. You send the NPS. The scores get charted. Somewhere in the open-text field is a detailed, specific, genuinely useful complaint from someone who is about to leave. Unless that person is suing you or defaming you publicly, the comment appears in a deck exactly once and is never seen again.

She points out that the same instinct explains a behavior everybody mocks and nobody stops doing, which is brands piling onto every trend within 48 hours of it breaking. Teams doing it may not even know why. The reason is that a trend is the safest possible risk. Everyone else is already there, it signals that you’re paying attention, and no one gets fired for it. A brand that jumps on a trend is a brand that would never have done that thing if it weren’t already trending.

Christina wants both halves, and she’s honest about why the pairing appeals to her. She was an English major who then taught math. Qualitative and quantitative are the same argument in 2 different accents. Cut one of them out and here’s what you get:

“What happens when you put creativity in a box? It stops being creative or it breaks the box and you have like a mutiny on your hands.”

Anyone who has worked inside that box knows the feeling. You get told you can color outside the lines on this campaign, and then you have to be really strict on that one, and nobody can tell you the rule that decides which is which. So you stop pitching the interesting version. That’s the quiet cost, and it never shows up in a test result.

The most expensive version of this runs every February. Super Bowl ads underperform what they could be, she argues, because of risk aversion at the buying level. A famous face buys you measurable recall. People remember the commercial and they remember who was in it. Whether any of it connects to what the brand actually stands for is a separate question that never gets asked, because the celebrity number is defensible and the brand argument is not.

Her reframe is what makes this usable. She wants teams to treat A/B testing as the cheapest permission slip they have rather than as a verdict machine.

“Think of it as a way to, this is how we can experiment. This is how we can play in very small places to see where we could go bigger.”

Phil made the same point from the analytics side. Test design orthodoxy says change one variable at a time, because if you change 3 things in an email you have no idea which one moved the number. That’s correct and it’s also a very small conversation. At the portfolio level, A/B testing is how you de-risk a big bet: run the weird idea in one geo, see what happens, and multiply it if it holds. The discipline of testing is what buys you the room to try something strange, as long as somebody is willing to propose the strange thing.

Experimentation programs get sold to executives as risk reduction, and that framing is exactly why so many of them produce safe, forgettable work. A program built to avoid being wrong will optimize its way to the local maximum and stay there for years. The teams getting real returns from testing are using it as cover for ambition, which requires somebody senior to say out loud that a flat test on an interesting idea is a better outcome than a 2% lift on a boring one.

Key takeaway: Pull the 5 most detailed negative open-text responses from your last survey and paste them into the next creative brief, unedited. Give the creative team the messy version instead of the score. Then pick one genuinely uncomfortable idea per quarter and run it as a contained test in a single market, with the explicit goal of learning rather than lifting a number.

Back to the top ⬆️

How Brands Build Core Memories Customers Cannot Buy

Brand teams measure recall aided, unaided, by first mention and by share of voice. It’s the metric the whole discipline was built on, and it asks a single question: when I say the category, does your brand show up? Christina borrows a better target from Pixar. In the film Inside Out, a core memory is the small number of moments that end up shaping who a person becomes. Psychologists call the real version autobiographical memory, and it does something recall never will. It becomes part of the story you tell about yourself.

A brand can be inside that story on purpose. Christina’s example is ESPN, and the mechanics are almost embarrassingly simple.

“You have more than recall. You have your fingerprints on their life.”

Here’s what they do. A crew goes up to the nosebleeds at a football or baseball game and finds somebody sitting as far from the field as the stadium allows. They usually pick a parent with a kid, which is the smartest part of the whole thing, because the emotional payload doubles when a child is watching. The crew asks how they’re liking their seats. Would it be cooler with a better view? The person has no idea where this is going. Then they get moved down. Sometimes they end up on the field, sometimes meeting the team, sometimes just close enough to see faces, with a bit of merch on the way out.

The moment sticks somewhere permanent. And the return on it arrives later, in the least glamorous circumstance imaginable.

Some Tuesday night the app won’t load, or the stream stalls in the 4th quarter, or the account page throws an error. That person hits the same friction everyone else does. What they have that other customers don’t is a once-in-a-lifetime day with the brand’s name attached to it. The irritation has to fight something, and it usually loses.

Christina draws a hard line between that and the thing most brands are chasing, which is attention. Attention isn’t scarce and it isn’t hard to get.

“Anyone can get attention. Train wrecks get attention, car crashes, attention, flooding, attention.”

The bar she sets instead is transformation, and she finds it with a purchasing test. Ask what your brand could give somebody that has no price attached to it anywhere. A discount fails that test. A gift card fails it. Faster shipping fails it. Being on the field passes, because there’s no checkout page for it.

She has a second example that shows how little of this requires a budget line. A couple got married and, as a joke, invited Bad Bunny to the wedding, fully expecting nothing. His team called back. They couldn’t come, but they were doing a thing, and would the couple be interested? The couple ended up married with Bad Bunny performing. That is a story those 2 people will be telling for the rest of their lives, and there was never a version of it available for purchase.

What that buys a brand outlasts every campaign it will ever run. When someone tells their kids and eventually their grandkids about the day they got to go down to the field, your name is in the sentence. You’ve stopped being a vendor they remember and started being a detail in their biography.

“You are seminal to who they are as a person. You can’t pay for that. That is bigger than any ad anywhere. It’s priceless.”

Brand measurement is built almost entirely around recall, which is the cheapest thing a memory can do. The industry has spent 30 years getting very good at buying the shallow end and has almost no shared vocabulary for the deep end, which is why the budget for a core memory program has to be smuggled in under experiential or community or advocacy every single time. Whoever builds the measurement layer for this is going to move a lot of money.

Key takeaway: Write down 3 things your company could give a customer that they couldn’t buy from you at any price. Access to a person, a stage, a credit, a decision, a room they’d never otherwise get into. Cross off anything with a dollar equivalent, because a customer can always buy that somewhere else. Whatever survives is the only raw material you have for building a memory, so pick one and give it to 10 customers this quarter.

Back to the top ⬆️

How B2B Brands Build Core Memories Without a Budget

There’s a reflex every B2B marketer has when this subject comes up. Sure, a stadium can put a kid on the field. We sell workflow software to mid-market operations teams. Nothing we ship is going to change anybody’s life. Christina’s response to that arrives before the sentence is even finished, and it’s 3 words long: yes, it is.

She has receipts from both sides of the budget question. The expensive version came from her time at HubSpot, where she created the Inbound Correspondence Program. The idea was to bring the company’s top advocates and biggest customer fans to Inbound and treat them the way the company would treat influencers. They’d create content, share it from their own perspective, and get real access to the event. 2 of them ended up hosting the main stage. There is no tier of a customer marketing program you can buy that produces that, and no competitor could hand it to them either.

“If you are the brand that helps someone get a job, they are gonna yell your name at every single vendor meeting.”

That quote comes from her second example, and it’s the one worth stealing, because it costs almost nothing and almost nobody is doing it.

Look at who has disappeared from your advocacy program in the last 2 years. A meaningful chunk of your best customer fans are no longer counted as customer fans, because the job they held when they loved you got eliminated. Most community programs handle this by quietly letting those members go dark. They can’t influence a renewal, they can’t attend as a customer, and they stop showing up in the reporting.

Christina’s move is to build them a room instead. A dedicated group inside your existing community, specifically for members who’ve been laid off and are looking for work. You’re saying the quiet part out loud to them: we know you can’t champion us right now because you’re not technically working, and we want to help anyway. If the community already exists, the space itself is free. What goes in it is mostly other people’s generosity:

• A place for members to post open roles they hear about.
• Shared resources and free certifications people find while they’re looking.
• Peers in the same position, because commiserating with someone who gets it is half of surviving a job hunt.
• An hour a month from your head of HR, talking through the hiring trends they’re seeing and the skills they’re screening for.

That last one is the whole thesis in miniature. It’s 60 minutes of one employee’s calendar. The audience is already assembled. And for the person who lands a role out of it, the sentence they carry forever is “I got this job because I was part of that community.”

Christina is blunt about where the money actually goes. Community always costs something, so the real question is which currency you’re spending. Teams default to thinking in budget, and budget is the one thing a lot of them don’t have.

“If you don’t have the budget for something, you do have time, and you do have access.”

The payoff isn’t subtle. Someone you helped through the worst professional stretch of their life will talk to procurement longer than anyone else on the call. They’ll bring you up in Pavilion, in closed Slack groups, in every vendor conversation where they have a vote. Her framing of what they’re saying: “They had my back. They helped me when I couldn’t help them.” You cannot generate that sentence with swag.

Which is where she loses patience with how the industry talks about this. Surprise and delight has become the reflex answer for any question about customer emotion, and it’s a fundamentals problem dressed up as a gifting problem.

“We just think surprise and delight. No. Go back. You gotta get the basics done first. What would transform someone’s life?”

Phil ran into the same wall from the content side. He messages every new listener who subscribes on LinkedIn and asks what’s keeping them up at night, mostly to plan future episodes. Earlier in the year the answers stopped being about martech. People were telling him to read the room, because they weren’t working on anything at all. Talking about customer data platforms and multi-touch attribution started to feel silly when a good chunk of the audience was out of work. So the show ran a series on the job market, brought on people who had recently been laid off so they could commiserate on air, and started shouting out unemployed listeners in the newsletter. Phil says one of them found a job out of it.

B2B loyalty budgets are allocated almost entirely to people who are currently in a position to sign something, which is a reasonable-sounding rule that costs companies their best advocates at the exact moment those advocates are most reachable and most likely to remember who showed up. The advocacy programs that will look smart in 3 years are the ones that track the person rather than the seat.

Key takeaway: Open a space in your existing community for members who are between jobs, and staff it with time instead of money. Book a recurring monthly hour with your head of HR to talk hiring trends and in-demand skills, and let members post roles and resources in between. Then stop removing laid-off members from your advocacy lists, because the person you help this quarter is the one arguing for you in a procurement meeting 2 years from now.

Back to the top ⬆️

What to Do When Transactional, Behavioral and Feedback Data Disagree

Anyone who has worked with customer data has run all 3 of these systems at once. Transactional tells you what someone bought and where they sit in the journey. Behavioral tells you what they did yesterday, what they’re doing now, and which segment they’ve fallen into. Feedback tells you how they interpreted any of it. The standard advice is to integrate all 3, because a single stream on its own can’t tell you what to do next.

Nobody tells you what to do when the 3 streams contradict each other. That happens constantly, and picking a winner by gut is how most teams handle it.

Christina’s answer starts somewhere unexpected. A disagreement between your data sources is itself a finding, because usually one of the streams has a reason to be wrong.

“I think you have the story. I think you have an unreliable narrator.”

Here’s the shape it takes in a real program. You have an advocate who genuinely loves the company. Their transactional and behavioral data says they’re struggling badly. Usage is down, tickets are up, they aren’t getting what they need. Their feedback says everything’s fine. It’s okay. It’s good.

Most dashboards will average that into a shrug. Christina asks a different question, which is whether this person has psychological safety with you. Because she’s had advocates do exactly this, and she knows why they do it. The fear is specific: if I tell Christina I’m unhappy, she’ll stop giving me opportunities and decide I’m a bad advocate. Acting fine is in their interest. So they act fine, the feedback stream lies, and the data conflict is the only evidence you get.

The repair is a conversation rather than a better model. You reach out, you tell them you saw something, and you make the terms explicit: this is a safe place, is there anything I can do? What she actually says to people is worth copying almost word for word.

“No. You still get opportunities. I’m not gonna break up with you because you tell me you’re mad. We’re fine, but I need to know. I can’t fix it if I don’t know what’s going on.”

Once the safety is real, the feedback stream starts telling the truth and the 3 sources line up again. The conflict was doing its job the whole time. It was showing a relationship problem that no amount of survey design was going to surface.

Why Your Most At-Risk Customers Never Answer the Survey

Christina has a love-hate relationship with NPS. She thinks it serves a real purpose and she’ll say so. She also says the picture it produces is never full enough to stake anything serious on.

“I would never bet my life on NPS.”

The structural flaw is a sampling problem hiding inside a sentiment metric. The customers closest to churning are frequently absent from the response set entirely, because they gave up on you before the survey landed. Their internal monologue is easy to reconstruct. Why would I spend 5 minutes on this? You’re not going to read it. You’re not going to fix the thing I want fixed. You’re not going to reply. What is the point?

So the people whose feedback you need most are missing from the one instrument built to collect it. You can see them leaving in the behavioral data. You can sometimes find them on a third-party review site describing exactly what went wrong, in detail they would never say to your face. The score on your dashboard stays healthy the entire time.

Her answer is proximity. Get as close to the customer as the business allows, run social listening properly, and treat the places customers talk without you as primary sources.

Why CX and Brand Are the Same Job

All of this leads her to a claim she flags as a hot take, and it’s the sharpest thing she says all episode. Support and social media are looking at the same customer problems from 2 different rooms. Support sees it in a closed, internal, ticketed space. Social, in her words, is a wild kingdom where you’re yelling at everybody and whoever sees it, sees it. They’re watching the same stories play out and they never compare notes.

Then she goes further.

“CX and brand, I would argue, this is my hot take, CX and brand is the same job.”

Test it yourself. Ask a brand person to define brand and ask a CX person to define CX, and you’ll get the same paragraph back twice. How people feel about you. The emotionality of it. What comes to mind when they see the logo or think about the product. Why they pick you over the alternative. 2 job families, 2 budgets, 2 sets of quarterly goals, and one definition. They never talk to each other, and Christina’s verdict on that is just: it’s wild.

Contradictory customer data is treated as a data quality ticket in most organizations, which is why it gets routed to the people least able to solve it. A conflict between what someone does and what someone says is a signal about the relationship, and relationships are not owned by the data team. Companies that learn to read disagreement as evidence rather than noise will find their churn list months earlier than the ones still waiting for a cleaner score.

Key takeaway: Write a rule into your customer health scoring that treats a conflict between behavioral data and self-reported feedback as its own alert, instead of averaging the 2 into a middling score. Route every one of those accounts to a live conversation, and open it by telling the customer their answers stay separate from their status with you. Then check how many of your last 20 churned accounts ever responded to a survey at all.

Back to the top ⬆️

Why Spotify Wrapped Works and What Gen AI Took Away

Spotify Wrapped became the reference every personalization deck reaches for, and it earned that by stacking things most brands attempt separately: listening history, visual creative, identity, social distribution, a personalized product, and a community moment, all firing at the same time in the same week. Then everyone else built their own version, most of them mildly embarrassing, and the format got diluted by imitation.

The interesting failure came from Spotify itself. In 2024, Wrapped went heavy on generative AI, including AI hosts reading users their own stats and riffing on their top genres. Users called it lazy. Adweek covered the company walking it back for 2025 and making the whole thing more human again.

Notice what nobody complained about. The numbers were correct. The data was as good as it had ever been. What people said was that it had no soul, which is a difficult thing to put in a post-mortem.

Christina’s take on why the original works starts with what people are actually doing when they post it.

“When I share my Spotify Wrapped, I am not just sharing this is what I listen to. This is a TLDR of who I am.”

She puts it in the same family as your Hogwarts house, your Myers-Briggs, your favorite book. All of those exist to serve a psychological need that has nothing to do with the underlying content. They’re shorthand you hand somebody so they can place you quickly. Wrapped happens to be the version built out of behavior you didn’t perform for anyone, which is what makes it feel earned.

And because it’s music, it carries things a purchase history never could. Look at the monthly breakdown and you can see your own year in it. Christina describes the experience of scrolling to March, seeing what she had on repeat, and thinking, I was going through it. Then a few months later, apparently I was in a better head space. Your listening data knows things about you that you did not write down anywhere.

So when a brand reads that back to you well, it feels like recognition. She calls it a very un-cringey, unpredatory way of saying this is who you are based on how you used our product. And people share it because being seen and heard is the entire reason most of the internet exists.

“Let’s be honest, we’re all a little bit narcissist. Social media exists because we wanna be seen and heard. If we didn’t have that innate need to be seen and heard, no one’s doing podcasts, no one’s doing video content.”

That explains the 2024 backlash precisely. A mechanical version of a self-portrait strips the emotionality out and leaves the accuracy behind, and the feeling that replaces recognition is closer to insult. You open the thing that’s supposed to be you, and your reaction is that this could not be less like me. The data was right and the mirror was wrong.

Her broader rule is worth writing on a wall somewhere: wherever there’s emotion in the experience, people want humans. Music is at the far end of that spectrum. In her words, it’s the human spark personified, so an AI narrator standing between her and her own year negates the thing that made it special. Spotify got back to the heart of it and the complaints stopped.

What the AI Avatar Pitch Misses About Why People Show Up

Phil hit the consumer version of this at his kitchen table. His dad started getting served ads for a podcasting tool that turns you into a digital avatar, sold on the premise that podcasting is heading toward AI avatars interviewing other AI avatars. His dad wanted to know if that was really where this was going. Phil’s answer was short: Dad, my show is called Humans of Martech.

“I am allergic to this.”

He’s also honest about his own use of the tools, which is where the line gets interesting for anyone running content operations. The research is AI-assisted. The questions get sharpened with AI. The newsletter and the blog post get repurposed with AI. He’s fine with all of it, because the story underneath is still a human one and the speed is the only thing being bought. The follow-up questions and the live dialogue in the room are the part no model produced.

Take the humans out completely and you have a different product that needs a different name. As he put it, at that point you’d have to call it Robots of Martech.

There’s a practical lesson buried in the Wrapped story that most teams are about to relearn the expensive way. Generative AI is being deployed hardest at exactly the moments where customers are most alert to whether a person was involved, because those are the moments with the highest engagement numbers attached. The emotional peaks in a customer journey are the worst possible place to save money, and they are the first place the efficiency case gets made.

Key takeaway: Map the 3 or 4 moments in your customer journey where people share something with their own audience, and mark those as no-go zones for generated content. Wrapped-style moments work because the customer is putting their identity on the line in public, so anything that reads as machine-authored costs them socially. Keep AI in the assembly and the analysis, and keep a human on the voice.

Back to the top ⬆️

Speaking of personalization and meeting people where they are… are you still operating like it’s 2013 and you haven’t updated your chatbot? We recently started collaborating with the Docket team and what they are building with their Inbound Demand Agent is super cool. Check it out.

Why Personalization Feels Predatory Even When You Have Consent

Consent has quietly become the whole compliance story in martech. If the banner fired, if the checkbox got ticked, if the DPA is signed, the data is fair game and the campaign ships. Everyone in the chain is doing their job correctly and the customer still ends up feeling watched.

Christina’s second book is built on the customer trust equation she developed, and it exists because the conversation about trust keeps stalling at the same place. Everybody agrees trust matters. She wants people to move past that soundbite on a panel and get to the how. Her contribution to the data half of that question is a concept she calls earned context.

“You could have a lot of data on me, either from brokers or because I’ve given you consent or whatever that is, but if you don’t have the earned context for it, you don’t have the relationship with me that means that it warrants you to have that data.”

The distinction is doing real work. Possession of data and permission to use it are legal questions. Earned context is a relationship question, and it asks whether the connection you have with this person warrants what you’re about to do with what you know. A brand you’ve had a 4-year relationship with can reference your history and it reads as attentiveness. A brand you’ve never bought from can reference the same behavior and it reads as surveillance, using data that came through channels you technically agreed to somewhere.

She’s clear-eyed about what the paperwork protects. It keeps you out of court. It does nothing else.

“The fine print alone will prevent you from me suing you, but it will not prevent me from being angry about it, from leaving, from telling everyone else on the planet not to use you.”

So the questions she wants asked before a personalization program ships are all upstream of the campaign brief. How are we getting this data? Where did it come from? What’s the context for the way we’re about to use it? Did we earn that context, or did we just acquire the record?

She gets a live demonstration of the gap every week. Her DMs are full of AI CX tools asking her to promote or endorse them. Her opening question back is always the same, which is how the product scales human contact, or how it replaces a human without the customer feeling the loss. 9 times out of 10 the answer she gets is that the tool personalizes everything.

“I just need to scream from the rooftops that personalization is not connection. They are not the same.”

That gap gets its own chapter in the second book. The working test she uses is short enough to run in your head. Data used with earned context produces the reaction “oh, you see me.” The same data used without it produces “oh, you are stalking me.” It’s the same record, the same accuracy, the same tooling. The line between the 2 reactions is thin and customers find it instantly.

Phil’s version of the fix comes from the privacy side, and it’s the most boring, most effective idea in the discipline. Data minimization inverts the default. Most martech teams collect everything on the theory that some future campaign might want it, store it indefinitely, write a vague sentence about deletion into the privacy policy, and figure out the use case later. Minimization asks what you actually need in the next 6 to 12 months and what you’re going to do with it. Anything that fails that question stays uncollected.

The follow-on question he likes is even simpler. What’s the worst thing that happens if you don’t track it? Most of the time the honest answer is nothing, and you’ve just removed a row from the breach you haven’t had yet. From there you get into third-party tags, piggyback tags riding on top of those, and privacy policies that describe something different from what the consent banner is actually doing. US teams are still catching up to Europe and Canada on most of it.

Privacy work in martech has been framed as a legal cost center for a decade, which is why it gets budget only after something breaks. Earned context reframes the same work as a retention lever, and that’s the version a CMO can actually fund. The companies that figure out how to say no to available data are going to have a durability advantage over the ones that treated consent as a green light.

Key takeaway: Add an earned-context check to your personalization review, right beside the consent check. For every data point a campaign uses, write down where it came from and what in the relationship gives you standing to reference it out loud. Kill any use where the honest answer is that a broker sold it to you or a banner technically covered it. Then run the same audit on the fields you collect and delete anything with no named use in the next 12 months.

Back to the top ⬆️

Where Influence Ends and Manipulation Starts in Marketing

Chapter one of Christina’s book goes deep into the machinery of persuasion, covering behavioral science, neuromarketing, eye tracking, biometrics, emotion and memory and reward, and sensory cues. Then it arrives at scarcity and urgency with a condition attached, which is that none of it should undermine long-term trust.

That condition is where the real argument lives, because low-stock counters and countdown timers are standard issue in D2C now and a good share of them are lying. Imagine the CEO who reads a case study about a countdown timer lifting conversion by some number and wants one on the site by Friday, with no actual inventory data behind it. Where exactly is the line?

Christina has a name for the thing on the wrong side of it. She calls it Machiavellian marketing, because the ends justify the means and everybody involved knows it.

“It also shows a contempt for your audience that you’re basically conditioning them to act feral and be mean and cruel.”

Her examples are recognizable to anyone who works in email or social. There was a send late last year that used a fraud alert as the subject line to drive opens, which she notes runs against CAN-SPAM, and which somebody in that building approved. Then there’s the whole genre of campaign built to anger people deliberately, where the anger is the media plan.

What bothers her about it is how easy it is. Manipulating people into rage takes almost no skill, and the attention economy has bent every incentive toward doing it. Everything became a tactic for attention. Everyone started thinking in short-term gains, counting views and mentions, and the old line about all publicity being good publicity quietly turned into the actual definition of success. There’s no accountability inside that definition, because negativity stops registering as a cost at all. Everyone’s talking about us, so it’s fine.

Her response is flat refusal. If that’s what winning means, she doesn’t want it, because what it does to the audience is the product.

Why Rage Bait and Goodwill Run on the Same Machinery

The behavioral psychology that makes rage bait work is the same behavioral psychology that makes goodwill work, and that’s what turns this from a complaint into an argument. Executives understand the mechanism. They’re choosing.

“Rage bait works the same way that goodwill works.”

She points to the World Cup and to Artemis as monocultural moments from this year that produced enormous attention and brand recall while running entirely on the positive side. The reach was there. The recall was there. The mechanism was identical. What changed was which emotion got farmed.

Her running example of the manipulative version is the ranked list. Rolling Stone puts out a top 100 vocalists and Celine Dion is missing, or Chris Cornell shows up at number 40, and the internet detonates. WatchMojo does the same thing. She likes WatchMojo, and she’s clear that there is always at least one placement nobody at that company genuinely believed in. The wrong entry is the product. You’ll keep watching and you’ll rage, and both of those are billable.

Her image for what this does at scale comes from 28 Days Later, where the monkeys are deliberately infected with rage so that anger is all they have left. Her take on the industry is that we’ve been running the same experiment on customers, saying we want attention at any cost, and if the angry version of you is the one who talks about us, that’s the version we’ll take.

The bill for that is arriving. Trust is down across every sector, and she thinks the deviation from basic marketing principles is a big part of why. The exhaustion people describe when they talk about wanting to stop doomscrolling, or wanting off social entirely, is the same exhaustion, and marketers helped manufacture it.

“We’ve manipulated the worst feelings out of us, and I don’t wanna see it anymore. No more dark patterns, as they say in UX.”

The usable line, then, has less to do with which technique you picked and more to do with what the technique requires the customer to believe. A countdown timer over real inventory is information. The same timer over invented inventory requires the customer to be wrong for it to work, and every tactic with that property is borrowing against a balance the brand cannot see. Sector-wide trust decline is what the aggregate bill looks like, and no individual campaign will ever show it on a dashboard.

Key takeaway: Run every urgency and scarcity element on your site through one test: does this stop working if the customer knows how it’s built? A countdown tied to a real deadline survives that. A low-stock counter with no inventory feed behind it does not, and it needs to come down. Do the same audit on subject lines, and delete any that borrow authority they haven’t earned, starting with anything that mimics a security or fraud notification.

Back to the top ⬆️

Why Audiences Are Going Back to the Humanities in the AI Era

The closing question on every episode is about energy. How do you decide what deserves it, and what keeps you aligned with the things that actually make you happy? Christina is a fractional chief customer officer, an author, a trust theorist, a community builder, a former teacher, a serious movie person, an avid reader, and a mother of 2. There is a lot competing for the same hours.

She answered by talking about what she’s been drawn to lately, which turns out to be the same answer.

“I think right now with AI and everything else, we are going back to the humanities. We are going back to our stories. We are going back to what makes us human.”

Her entry point was Spider-Man: Brand New Day, which she saw along with everyone else. What she was looking for in it was hope and responsibility, and a hero who is vulnerable, who carries sadness and loneliness the way the rest of us do. Tom Holland is absurdly talented and Spider-Man has been beloved for decades, and she thinks the size of the response comes from somewhere else.

What people are buying tickets for is the odyssey. A person going through it who is going to make it home, and who is going to deal with the people who should never have been in his house. You go and you watch a guy who is genuinely struggling, and you sit there hoping he wins and finds himself by the end of it. That’s an old shape. It’s the oldest one there is.

Christina is unembarrassed about loving the canon. She reads constantly, she loves the hero’s story, and she connects it straight back to the work: voice of the customer is storytelling, and storytelling is the whole game.

Her evidence is just the market. Look at what’s doing numbers at the box office, at the books moving right now, at the shows people are actually finishing.

“People want something to believe in. They want to see people that are human, that are flawed, that are trying to get through their day, that are trying to make the world a better place.”

Her closing note is optimistic in a way that’s rare in a conversation that spent real time on rage bait and dark patterns. There are signals everywhere, she says, that people want a better world. The demand side is already voting for it. Whether brands notice is a separate question.

Marketing has spent a decade hiring for the technical end of the funnel and treating the humanities as a soft skill you pick up on the side. The people who can build the model are now plentiful and the people who can tell you what story the model should serve are not. That imbalance is going to correct, and it will correct through hiring long before anyone writes a think piece about it.

Key takeaway: Take the campaign you’re building right now and name the person it’s about, then say out loud what they want and what’s in their way. If you can’t do it in 2 sentences, you have a targeting brief instead of a story and the creative will come out flat. Make that the first gate on every brief, ahead of the audience definition and the channel plan.

Back to the top ⬆️

Episode Recap

Christina Garnett makes one central argument across this conversation: martech got extremely good at accuracy and never built the muscle for meaning, and customers have already priced that in. Every capability the industry treats as personalization, from merge fields to behavioral triggers to lookalike targeting, is now available to anyone with a credit card. What that produced is a generation of campaigns that are mechanically correct and emotionally flat, and an audience that can recognize the pattern without ever having read a marketing book.

The tactical thread running under the chapters is a repeated instruction to go looking for the thing your dashboard cannot show you. Qualitative data gets ignored because nuance is uncomfortable, so the negative open-text response appears in a deck once and dies there. Your most at-risk customers never answer the NPS at all, because they gave up on you before it arrived. An advocate whose behavioral data screams while their feedback says everything’s fine is an unreliable narrator, and the only instrument that fixes that is a conversation with enough psychological safety in it that the truth costs them nothing. In every case the signal is available. It just lives outside the system built to measure it.

On the creative side she wants marketers to stop treating experimentation as a verdict machine and start using it as cover for ambition, because a testing program built to avoid being wrong will optimize itself into the local maximum and stay there. The alternative she offers is core memories, the moments a customer could not have purchased from you at any price. ESPN moving a family from the nosebleeds to the field is the consumer version. The B2B version is cheaper and almost nobody runs it: a space inside your existing community for members who have been laid off, staffed with an hour a month of your head of HR’s time. The person who finds work through it will argue for you in procurement meetings for years.

The ethical spine of the episode is earned context, the concept from her second book. Consent and possession are legal questions with legal answers. Earned context asks whether the relationship you have with a person gives you standing to use what you know about them, and it explains why 2 brands can run the identical personalization play and get “oh, you see me” from one and “oh, you are stalking me” from the other. She extends the same logic to influence tactics, calling the manipulative end of it Machiavellian marketing, and pointing out that rage bait and goodwill run on identical behavioral machinery. The choice between them is a choice, made by people who understand exactly what they’re doing.

She leaves a few tensions open rather than resolving them. Personalization is not connection, which is a clean line to say and a hard one to operationalize when your board is asking for pipeline this quarter. Her hot take that CX and brand are the same job is almost certainly correct and has no obvious implementation path in a company where they report to different executives with different targets. And her closing optimism, that the box office and the bestseller lists are signalling a return to the humanities, sits right next to her own account of an attention economy that keeps rewarding the opposite. Both things are happening at once, and marketers get to pick which one they staff for.

Connect with Christina on LinkedIn, and find Transforming Customer-Brand Relationships wherever you buy books.

Full episode ⬇️ or Back to the top ⬆️

✌️

—
Intro music by Wowa via Unminus
Cover art created with Midjourney (check out how)

Apple •Spotify• Pocket Casts •Youtube •Overcast •RSS

Related tags

See all episodes

Future-proofing the humans behind the tech

Apple •Pocket Casts•Google •Overcast •Spotify •Breaker •Castro •RSS

• LinkedIn
• Twitter
• Instagram
• YouTube
• RSS Feed