Humans of Martech

What's up everyone, today we have the pleasure of sitting down with Alexey Komissarouk, Member of the Technical Staff at Anthropic.(00:00) - Intro (01:22) - In This Episode (04:44) - How Engineers Actually End Up Specializing in Martech (08:47) - Security Approval Is the Real Bottleneck at Anthropic (14:34) - Why Anthropic Builds Its Own CDP Instead of Buying One (16:50) - How the Vibe-Code Era Is Rewriting Build Versus Buy Math (22:18) - Who Owns a Vibe-Coded App After Its Builder Leaves (28:25) - Why the Data Warehouse Beats Every Other Part of Your Martech Stack (37:49) - Which Martech Tools Survive Even Though You Could Replace Them (41:59) - How to Avoid Drowning in Shadow Martech Tools (48:07) - What Happens to Marketers When Claude Can Just Do the Work (51:36) - How Alexey Decides What Deserves His Energy Summary: Alexey Komissarouk builds martech infrastructure at the company with the most reason to trust AI to build it for them, and the picture he paints is refreshingly unglamorous. Terraform pull requests, a data warehouse doing more work than any vendor, and a security review process that's genuinely harder than the code itself. He explains why Anthropic still keeps Linear around, why the company builds its own CDP instead of renting one, and why an old thesis about IKEA shelves needed a serious rewrite. He also admits he doesn't actually know who owns half the internal tools people have built inside Anthropic, and he's not alone in that. If you've ever wondered what build versus buy looks like once building costs almost nothing, this is the episode that answers it.About Alexey KomissaroukAlexey Komissarouk leads growth engineering and martech at Anthropic, where he's spent the past year and a half rethinking which tools the company builds versus buys as AI changes the economics of both. He previously led Growth Engineering at MasterClass from 2020 to 2023 and spent four years at Opendoor as an early engineer and later a Growth Engineering manager. Before that, he ran a product engineering consulting practice for companies including Dropbox, Pebble, and Binti, and in 2014 he left San Francisco to co-found Hacker Paradise, a work-and-travel company that has since run trips to more than a dozen locations for over 800 alumni. He graduated from the University of Pennsylvania with a computer science degree, where he started the PennApps hackathon.Why the Best Martech Engineers Never Planned to SpecializeAsk most engineers what they want to work on, and martech rarely makes the list. It carries a reputation problem, unglamorous and marketing-adjacent, the kind of work that gets assigned rather than chosen. Alexey built an entire career on being the exception, and the path there had nothing to do with a plan.He grew up in Israel reading Joel Spolsky, came to the US for college, and started a company right out of school that ended in a messy founder breakup. A stretch of consulting followed, along with a bootstrapped travel company, Hacker Paradise. Then came the moment every founder-adjacent engineer eventually hits: watching former classmates land at Google and Facebook while he was, in his own words, futzing around for half a decade. He needed something that paid well and still felt like solving a real problem. That search led him to growth.Martech specifically was never the plan. No engineer, Alexey says, wakes up excited to work on it. It sounds boring until you've felt the pain yourself, and he had. At his own startups, he was the one setting up email automations, and he remembers exactly what it felt like to get pulled off real work because a marketer needed a template changed in code. That irritation is what turned into a career. Opendoor hired him to rebuild their long-term email sequence, a problem he already understood cold from years of iterating on Hacker Paradise's own systems.He ran that function at Opendoor, took it over again at Masterclass, and now runs it a third time at Anthropic. He describes himself as the engineer who never minded doing the work everyone else avoided, and that pattern has built the same career three times over.The engineers who end up owning martech well are rarely the ones who set out to specialize in it. They're the ones who got annoyed enough by a blocked marketer that fixing the problem became more interesting than avoiding it.Key takeaway: Identify the recurring request that annoys your engineering team most, whether it's an email template edit or a landing page tweak, and treat it as a real problem worth solving instead of an interruption to survive. Build the tool that makes that specific annoyance disappear for good. The engineers who end up running martech well are usually the ones who fixed their own irritation first.Security Approval Is the Real Bottleneck at AnthropicEvery fast-moving company claims to move fast, right up until security gets involved. At Anthropic, that friction shows up in a specific place: the moment a marketer's vibe-coded app is ready to connect to something real.The upside is a real departure from a typical martech org. Marketers at Anthropic don't file a ticket and wait. They build the tool themselves, and martech's job shifts from building it to making sure it holds up and plays well with everything else. Alexey says that's been the pleasant surprise: instead of a request landing on his desk, a finished app does.Security is the part that moves slower. Anthropic builds technology that's explicitly dual use, and that changes the math on every internal tool. Add an MCP connector to a martech app, and that app becomes a genuine attack surface. If it gets compromised, someone could use it to reach Anthropic itself. Approvals that would have taken a day at a previous job, the kind of thing nobody would have thought twice about, now run through a real review process.Why Anthropic Still Buys Some VendorsPart of that scrutiny comes from who Anthropic sells to. It has enterprise customers in health tech and fintech, which means it inherits a lot of the same compliance obligations as a vendor to them. That context shapes how the team decides what to build versus buy.Security reviews get treated as an afterthought in a lot of engineering cultures, something to route around rather than respect. Anthropic's own product makes that posture harder to justify. When the thing you build is the reason attackers would want in, slowing down to check becomes part of the job rather than an obstacle to it.Key takeaway: Map every internal martech tool that connects to an external service through an API or an MCP connector, and treat each connection as a potential attack vector instead of a convenience. Route anything handling customer or model data through a real security review before it ships, even when building it only took an afternoon. The friction is real, but skipping the review is worse.Why Anthropic Builds Its Own CDP Instead of Buying OneLook at almost any martech category right now and you'll see the same thing happening. Vendors that started out solving one problem keep bolting on adjacent features until they all start to look like the same product wearing a different logo. Alexey has a name for it.He speaks well of the category too. He's used Census and Hightouch at past jobs and as a consultant, and says he'd reach for one of those tools again at a growth-stage startup. Anthropic ended up building instead, mostly because so much of what the company needs is custom enough, and it has the headcount and engineering capacity to do it well.The Tool They Replaced: the CDPAsked what Anthropic replaced despite having a perfectly good vendor solution, Alexey doesn't hesitate.An...

What's up everyone, today we have the pleasure of sitting down with Lauren Stein, former Chief of Staff at GrowthLoop.Summary: Lauren Stein spent four years living the chief of staff job at GrowthLoop before most companies had a real name for it, and now that title is showing up explicitly in job postings recruiting straight out of marketing ops. This episode covers the scaffolding analogy for why growing companies actually need the role, why seeing the whole chessboard means learning to make decisions with data that's never fully clean, and why product management is becoming a real requirement for chief of staff hires. Lauren also gets into what happens to organizations once AI makes every individual faster but nobody's coordinating the pieces between them. If you've ever felt like your marketing ops seat puts you closer to the executive table than your title suggests, this one maps out exactly how to test that.About Lauren SteinLauren Stein spent four-plus years at GrowthLoop, rising from HR coordinator to operations lead to chief of staff, working directly with leadership on strategic planning, hiring, and company-wide execution. She built her career at the intersection of people, operations, and execution, managing cross-functional work that touched customers, partners, boards, and internal teams. Before GrowthLoop, she held roles at Contently, AdHawk, and Arkadium, building a foundation in people operations and startup environments. Today she works as an ops lead and head of ecosystem, known across the martech community as a networker, community builder, and startup mentor.Is Marketing Ops Already Preparing You To Be A Chief Of StaffMarketing ops job titles have been in flux for a while now. Marketing engineer, GTM systems owner, RevOps lead, the label keeps shifting depending on who's writing the job description. Underneath all of it sits the same tension: marketing ops people understand the systems, the data, the workflows, and the handoffs better than almost anyone in the building. They often know how the business actually runs better than finance does. And they still get treated like a downstream function, pulled in after the decisions that matter have already been made.Chief of staff roles are now explicitly recruiting from that exact pool. ServiceNow has posted openings looking for marketing ops and RevOps backgrounds. The pitch is straightforward: you already sit close to how revenue actually moves through an organization, so you're a step closer to the room where decisions get made, even if you haven't made one yet.Lauren spent four years watching that requirement shift in real time. The chief of staff hiring bar used to run through a specific pedigree: biz dev, consulting, investment banking. If you didn't have one of those three on your resume, you weren't getting a look. That bar has moved, not because the job got easier, but because the skill set that actually matters changed. A product manager who learned what the customer needed and what a sales team needed builds the same muscle a former banker used to bring, just from a different direction.That shift matters more for marketing ops and RevOps than for almost any other function, because those roles already sit on top of the thing every chief of staff eventually has to master: how the business actually drives revenue. You don't need to have made the strategic calls yet. You need to understand why they get made, and marketing ops people spend their whole job inside that logic. Lauren calls it auditioning for a role before you technically have it.The catch is that understanding the mechanics of revenue and being trusted to make trade-offs about it are two different skills. Marketing ops has the first one by default. The second one is what separates a systems person from an executive, and it's the entire subject of what comes next.Key takeaway: Audit your own role for revenue fluency before assuming you need a title change to get taken seriously. List every place you already have visibility into how decisions get made, who influences them, and what data feeds them, even if you're not the one making the call. That list is your existing case for a broader mandate, and it's stronger than most marketing ops people give themselves credit for.Chief Of Staff Is About Trade-Offs, Not Just Bigger ScopePut a chief of staff job description next to a marketing ops job description and they look almost identical from a distance. Both sit in the connective tissue between departments. Both deal with messy, cross-functional problems that nobody else wants to own. Both carry responsibility with no real authority to back it up. The overlap is real enough that plenty of marketing ops people could probably do a decent chunk of a chief of staff's job tomorrow without changing a single skill.What separates them comes down to what you're allowed to sacrifice, and how comfortable you are sacrificing it, not scope on paper.Marketing ops optimizes a function. Get the systems right and revenue moves, teams get unblocked, the machine hums. But fixing one function can break another without anyone noticing right away. A cleaner attribution model might starve a channel that was working for reasons the dashboard never captured. Better lead scoring might overload a sales team that isn't staffed to handle the new volume. Marketing ops rarely has the full picture to see that trade-off coming, and even when it does, it usually doesn't have the mandate to choose which function gets protected and which one absorbs the hit.A chief of staff's entire job is making that choice on purpose. Lauren describes it as thinking a few steps ahead: focus engineering on product this month, accept that another team falls behind temporarily, because the executive team already knows where the business needs to be in two years and can absorb the short-term cost. That's a different relationship with risk entirely, one where you're accountable for what you didn't fix as much as what you did, not a bigger version of marketing ops.Marketing ops without that context defaults to what Lauren calls the local maxima: optimize what's in front of you because that's the only thing you have visibility into. That's a structural limitation of the seat itself.Key takeaway: Before your next optimization project, map who else touches the outcome you're about to change. Ask marketing ops leaders one question before greenlighting a fix: what does this break somewhere else in the stack, and who owns that consequence? If you can't answer that question, you're optimizing for the local maxima instead of the business.Building Trust With ICs Is The Real Entry Point To Chief Of StaffAsk most people what a chief of staff actually does day to day, and the answer usually lives at the executive level: running OKRs, sitting in leadership meetings, prepping the CEO's board deck. That's real, but the job actually starts on the ground floor, with the individual contributors who have zero interest in a seat at the table and every interest in someone actually listening to them.Lauren's first move in any new chief of staff role has nothing to do with strategy decks. She meets with every single person on the team and asks the same handful of questions: what's important to you, what do you value, what should I focus on, where are the gaps, what shouldn't I do. When she joined GrowthLoop, the company was around fifteen people, so that meant every employee, no exceptions. Some had sharp answers to all five questions. Some had none. She met with them anyway.The logic underneath that ritual is simple. The person in front of a customer all day, or watching a deal fall apart in real time, is usually the first to spot what the business actually needs. If that observation never travels past their manager, the execu...

What's up everyone, today we have the pleasure of sitting down with Anna Gelfand, Director and Head of Lifecycle & Retention Marketing at Miro.(00:00) - Intro (01:14) - In This episode (04:29) - Why Lifecycle Marketing As We Know It Has No Future (08:53) - How To Codify The Qualitative Side Of Lifecycle Marketing (14:52) - Testing Email Campaigns On Digital Twins Of Your Customers (25:05) - Shipping The Criteria Instead Of The Campaign (34:23) - Why Lifecycle Teams Should Stop Calling Themselves The Email Team (42:32) - Why Your Workflow Should Define Your Martech Tools (49:12) - What Lifecycle Marketers Will Do When Agents Run Campaigns (51:58) - The Skills Lifecycle Marketers Should Double Down On In The AI Era (01:00:33) - How To Decide What Deserves Your Energy At Work (01:02:38) - Fiction Book Recommendations From A Marketing Leader Summary: Anna has a take she'll defend on stage: lifecycle marketing as we know it has no future. She breaks down what codification actually does to the job, why she built digital twins of 100 real customers to test campaigns without ever bothering a live user, and how the smartest lifecycle teams are quietly refusing to be called the email team. She gets into the 20/80 flip between building and thinking, why she'd rather own responsibility than a channel, and the one belief that survived 3 ESP migrations and 7 years at Miro. Stick around for the part where she admits she kinda likes suffering, intellectually speaking.About Anna GelfandAnna Gelfand is the Director and Head of Lifecycle & Retention Marketing at Miro, based in Amsterdam, where she's spent close to 7 years building and running lifecycle and retention for a global product-led company at real scale. She leads a team responsible for email, in-app, and the wider customer journey, and she's known for treating lifecycle as the connective brain of the organization rather than a single channel.Before Miro she worked in lifecycle and retention at Carrot Quest. She's also comfortable holding a contrarian view worth arguing over: lifecycle marketing as the industry knows it doesn't have a future, and its center of gravity is already moving toward strategy, content, and codified systems that agents run.Why Lifecycle Marketing As We Know It Has No FutureSpend enough years in lifecycle marketing and the work starts to rhyme. You build the segments, ship the campaign, wait for the numbers, and a large share of the time the test comes back saying nothing at all. Next quarter you run a version of the same thing with a slightly different subject line. When the uplift arrives, it's a fraction of a percent for months of effort.Anna hit that wall at Miro, and she's careful to point out it started before AI was disrupting everything left and right. She's been there a long time, long enough to watch the product and the market reinvent themselves more than once. By that point you've already picked the obvious wins, then the not-so-obvious ones, and what's left is weeks of work for a result that refuses to tell you anything.That line sounds strange until you've lived it. A clearly negative result kills an idea and hands you a lesson. An inconclusive one burns the same 3 weeks and teaches you nothing, so you're right back where you started with less time on the clock. Anna's take on the plateau is that manual, human-level optimization simply can't reach the relevancy customers deserve anymore. That's especially true at Miro, where the product is an open canvas and users can build almost anything on it. You know your personas and your core use cases, but hand-tuning content one segment at a time won't make an email feel genuinely built for the person opening it.Here's the part that keeps Anna in the craft anyway. Lifecycle is intimate, even when you're sending to millions and thinking in the billions. The human aspect of the communication is the whole point. So the real question she's chasing isn't whether to throw technology at the problem, because everyone will. It's what actually survives once you do. The plateau isn't a Miro problem or a tooling problem. Any mature lifecycle program eventually runs out of manual optimization road, and the teams still treating inconclusive tests as a normal cost of doing business are the ones that get lapped first.Key takeaway: Pull your last 2 quarters of lifecycle experiments and count how many returned a clear win or a clear loss. If most came back inconclusive, stop feeding new tests into the same tired playbook. Redirect that time toward a faster way to kill weak ideas before they eat a full test cycle.How To Codify The Qualitative Side Of Lifecycle MarketingCodification is the word doing the rounds right now, borrowed from Luke Kline's writing on the topic. The idea is simple enough: take the marketing knowledge locked in your head and your docs, make it accessible to AI agents, and let them operate on it directly. It works beautifully for the quantitative half of lifecycle. Propensity models, churn scores, who bought what and when, all of that is already data, so agents eat it easily. The harder question is the other half. How do you codify a gut feel, the sense of what makes a subject line bad, a nudge too pushy at the wrong stage, or a campaign a little too clickbaity?Anna's answer starts with a shrug and a strong opinion at the same time. Codification is inevitable, she says, because lifecycle was always the most technical, most data-informed corner of marketing. Expecting AI to leave that part alone would be strange. So the work becomes feeding agents the right tools and the right context.That tension runs through everything she describes. Anna trusts data and segmentation when the warehouse and the models are in good shape. She does not yet trust a machine to write the message. Her example makes the stakes concrete. Miro is a B2B2C company with a wall of self-serve users on one side and large enterprises on the other. She's always wanted lifecycle communications that feel like a partnership rather than a vendor broadcast, and doing that by hand only scales to the biggest accounts. Codification changes the math. Now an agent can pull a public customer's stated goals from their own reports, cross them with Salesforce and Gong data, and paint the email in that customer's colors with their use cases baked in. The account-based treatment she could only afford for a handful of logos becomes something she can run wide.The catch she keeps circling is judgment. She won't call it control, but she wants a clear moment where a human applies quality and qualitative feedback before anything ships. Part of that is buying back time to actually talk to users again, which lifecycle teams almost never do enough of. The other part is a bet on content. As AI floods every inbox with more messages than anyone can read, the scarce thing stops being the ability to produce and becomes the ability to stand out.The uncomfortable implication for most teams is that codifying the qualitative layer isn't a copywriting problem, it's a context problem. If your brand voice, your stage-by-stage rules, and your customer intelligence live only in senior people's heads, an agent has nothing to operate on and will happily generate confident garbage at scale.Key takeaway: Write down the judgment calls your team currently makes by instinct. Document what a bad subject line looks like, what's too pushy at each stage, and the brand rules you'd never let slip. Feed that to your agents as context, then set one named human checkpoint where quality gets approved before anything sends.Testing Email Campaigns On Digital Twins Of Your CustomersPlenty of marketers have built a digital twin of them...

What's up everyone, today we have the pleasure of sitting down with Melissa M. Reeve, author of Hyperadaptive: Rewiring to Become an AI-Native Organization and co-founder of the Agile Marketing Alliance.(00:00) - Intro (01:09) - In This Episode (04:33) - What Marketing Looks Like If AI Had Always Existed (08:47) - Why AI Is Easy to Use but Not Easy to Learn (15:34) - Why Marketing Ops Sits Where Friction and Fidelity Break (19:20) - How to Measure Your Own Coordination Tax (29:16) - The 5 Capabilities of an AI Native Organization (35:18) - Why AI Sensing Fails When You Measure the Wrong Things (40:00) - Where Ops Jobs Go When Anyone Can Create Fields (45:42) - What a Response Framework Has That a Dashboard Doesn't (52:24) - Why Marketing Needs a Learning Integration Manager (55:56) - A Human Shift Enabled by Technology (58:08) - Happiness, Satisfaction, and the Act of Noticing Summary: What would your marketing team look like if AI had always existed? Melissa spent 30 years studying how organizations actually change, from the floor of a Japanese truck factory to the front lines of agile marketing, and her answer starts with ripping out an org chart designed in 1911. Along the way she explains why "go play with AI" splits your team in 2, how to put a hard number on the coordination tax eating 60 to 70% of managers' time, and the 5 capabilities that separate AI native companies from companies with AI licenses. Stick around for the McFlurry machine that explains predictive sensing better than any martech deck ever has.About Melissa ReeveMelissa M. Reeve is the author of Hyperadaptive: Rewiring to Become an AI-Native Organization, published by Gene Kim's IT Revolution. She studied the Toyota Production System at Waseda University in Tokyo and learned it for real on the factory floor at Hino Motors, watching workers pull the andon cord that lets any one person stop the whole line. She went on to become the first VP of Marketing at Scaled Agile, the company behind SAFe, and co-founded the Agile Marketing Alliance, where she's been pioneering agile marketing since 2011. Hyperadaptive distills 3 decades of systems thinking into a playbook for rebuilding organizations around AI.What Marketing Looks Like If AI Had Always ExistedWhen factories swapped steam for electricity, almost nothing changed for decades. The buildings kept the same layout steam had forced on them, one big engine driving belts and pulleys through the floor, and the efficiency gains stayed embarrassingly small. The revolution only arrived when somebody asked what a factory would look like if electricity had always existed. That question is the spine of Melissa's book, with AI playing the role of electricity. Most marketing teams today are running a copilot bolted onto the old layout and wondering where the transformation went.Melissa's answer starts a century back. In 1911, Frederick Winslow Taylor invented scientific management, and with it the idea that there's one best way of doing something and that it's management's job to impose that way on everyone below them. Every marketing org chart still runs on it. You start as a junior copywriter, become a senior copywriter, then a director, then the head of content, and at each rung your job is to pass the one best way down the ladder. That structure is so embedded in how marketing works that most of us have never thought to question it.Now drop AI into that world. A system that can deploy an entire campaign at once, text, graphics, and programmatic advertising included, makes the whole ladder look strange. And history tells us exactly what happens to the people on it.Melissa has studied enough automation waves to trust the pattern, from factory floors to the washing machine. Nobody she's talked to wants to go back to beating laundry on rocks. In her AI native marketing org, an AI completes campaigns end to end while more humans babysit the automation. Somebody builds and rebuilds it as models change. Somebody watches for brand drift and data drift, and catches the agent before it ships its own OpenAI-and-Hugging-Face moment, the cheerful "oops, didn't mean for that to go out" post nobody approved.She's realistic about the timeline too. Reimagining the process is the easy part; the org also has to realign job roles, incentives, and the entire structure around them, and those things move slowly because Taylor's ladder is bolted into all of them. The teams that win will be the ones willing to redraw the factory floor, not the ones with the best copilot licenses.Key takeaway: Pick one campaign workflow and redesign it on paper as if AI had always existed, with no approval ladder inherited from the pre-AI org. Then list the new jobs that redesign creates: who builds the automation, who monitors brand and data drift, who rebuilds when the model changes. That list is your team's next set of roles, and drafting it now beats discovering it during a reorg.Why AI Is Easy to Use but Not Easy to LearnNovember 2022 hit marketing first and fast. Melissa had spent a decade bringing agile software principles into marketing, learned lean the hard way on the floor at Hino, a Japanese truck maker, watching workers pull the andon cord that lets any one person stop the whole production line, and was running the Agile Marketing Alliance when ChatGPT sucked all the oxygen out of the room. She calls that moment the great forcing function. Her first reaction was the same one you probably had: the ground shook a little, and a blog post got written awfully fast.What happened next is a journey map she's watched play out across the whole industry. Year one was polarizing. Some people geeked out and leaned in, the headlines screamed that all the jobs were going away, and the resisters dug in hard; she lost a friendship to a flat "don't bring it up, I'm anti-AI." Melissa herself spent 6 months in what she calls the confusion zone, in mild denial, until her business partner Jim Uhl told her to pay attention because this one was big. Then came the early wins era: enterprises bought licenses, handed them out, and delivered the 4 most dangerous words in corporate AI adoption, go play with AI.Anybody can walk up to a piano and bang on the keys. Becoming a concert pianist takes years of deliberate practice, and the people getting real results out of AI have put in exactly that kind of time learning what these systems can and can't do. That gap is why "go play with AI" felt to Melissa like screaming into the void. Play produces random acts of AI, scattered experiments that never compound. Worse, it produces what she calls the bifurcation problem: a handful of power users pulling away from everybody else, while the everybody-else group waits for support that never comes.By last year the chorus had moved on to agents, agents, agents. But teams still don't know how to get from here to there, and that gap is the reason her book spends most of its pages on the path itself: the systems you need to put in place, the named roles and named responsibilities that actually shift the mindset and the work. Factories weren't rewired overnight, and neither are marketing orgs. The rewiring happens incrementally or it doesn't happen at all.Key takeaway: Replace "go play with AI" with structured support before your team bifurcates into power users and everyone else. Name the roles responsible for AI adoption, give each experiment an owner and a goal, and build a shared place where learnings land so they compound instead of evaporating. Treat AI fluency like piano lessons: scheduled, deliberate, and cumulative, never left to whoever happens to noodle around on the keys.Why Marketing Ops Sits Where Friction and Fidelity Break

What's up everyone, today we have the pleasure of sitting down with Gary Amaral, marketing leader at Docket.(00:00) - Intro (01:15) - In This Episode (04:35) - What Lead Scoring Was Really Hired to Do (12:13) - Why Lead Scoring Models Turn Into Black Boxes (16:04) - What Is an Agent Qualified Lead (24:02) - How Agentic Qualification Goes Beyond AI Scoring (27:48) - What AI Agents Do That Drift Style Chatbots Never Could (34:35) - Why Buyers Tell AI Agents What They Hide From Sales Reps (40:05) - Who Needs an AI Agent on Their Website (42:48) - How to Keep the AQL From Becoming the Next Vanity Metric (50:11) - How to Introduce AQLs Without Breaking Your Forecasting (56:30) - Will AI Agents Replace Intent Data and Marketing Automation (59:41) - Why Work Life Balance Is the Wrong Goal Summary: Gary built lead scoring so ambitious it crashed a Marketo pod and took dozens of companies down with it, then co-founded a scoring company, sold it, and switched sides of the argument. Now he says the whole category rests on guesswork: scores infer what buyers want, while an AI agent can just ask them. He explains why every rep would trade 20 scored leads for one hand-raiser, how Docket's agent Aura lifted engagement 8X by moving into the homepage hero, and why he refuses to be measured on MQLs anywhere he works. Along the way you get a Waymo analogy, a Meat Loaf methodology, and a bet that agents will eventually replace your bloated marketing automation platform. By the end you'll know exactly how to pilot AQLs without giving your CFO a forecasting heart attack.About Gary AmaralGary Amaral has spent his career on the question lead scoring was invented to answer: which buyers actually deserve a sales rep's time. He chased it at BlackBerry, then at Hootsuite, where the demand center his team built included a scoring system so complex it took down an entire Marketo pod. That week of hand-routing leads became the genesis of Breadcrumbs, the lead scoring company he co-founded and later saw acquired by MadKudu.Today Gary leads marketing at Docket, the AI agent platform behind Aura and the agent qualified lead, where he is making the case that asking buyers what they want beats inferring it from their clicks.What Lead Scoring Was Really Hired to DoEvery marketing ops team has a scoring model somewhere in their instance that nobody fully trusts. Points for a webinar, points for a pricing page, a threshold someone picked in a meeting 3 years ago. Strip away the arithmetic and the model was only ever hired to answer one question: who should sales reach out to right now?Gary has been chasing that question since the BlackBerry days. By the time he landed at Hootsuite, lead scoring had grown into something much bigger, a full demand center with scoring at its core. And the system his team built grew so elaborate that it broke more than his own instance.He then spent a week manually sorting and routing leads while Marketo dug itself out. Imagine the tedium: a demand leader hand-triaging spreadsheet rows because the machine built to do it had flattened the neighborhood. That week became the genesis of Breadcrumbs, the scoring company he went on to co-found. His pitch was simple. He describes himself as "just an average guy" without a PhD in statistics, carrying years of experience feeding sales teams quality leads. Recency, frequency, fit, and intent, combined in a way that tells you where to send a lead and which seller should get it. Breadcrumbs set out to bring that sophistication to everyone else.Then the market moved. Gary names 3 headwinds that hit the company:AI-native competitors arrived, and if you squinted you could already see how agents would change qualification entirely, Legacy platforms finally invested, with HubSpot shipping scoring that was still not great but good enough for teams already paying for it, Consolidation swept the category as vendors hunted for efficiencies and expanded offeringsThat last headwind doubled as an exit, and MadKudu acquired Breadcrumbs. The pattern says something uncomfortable about martech point solutions: when your product's output is a number that feeds another team's queue, you are one platform release away from becoming a feature.Key takeaway: Audit your scoring model against the job it was hired to do. List the last 20 leads it routed to sales and ask a rep which ones deserved the outreach. If the rep's answer and the model's answer disagree on more than a handful, the model is scoring activity, and you should rebuild the weights around closed outcomes instead of page values someone negotiated in a meeting.What Made Advanced Lead Scoring Different From Point SystemsThe classic model runs on decree. Somebody decides the pricing page is worth 20 points and a blog post is worth 10, usually during a horse-trading session where sales insists that anyone who looks a certain way and clicks a certain thing must be hot. Marketing goes along with it because keeping the peace is cheaper than arguing. The weights encode office politics, and everyone quietly knows it.Advanced scoring, the kind Breadcrumbs built, flipped the direction of authority. Instead of a human declaring what matters, the model watched signals over time, tied them to actual outcomes, and surfaced the combinations of behavior that genuinely preceded a purchase. More data sources feeding the model, and math deciding the weights. Harder to explain, Gary admits, but honest in a way the point system never was.That last part deserves more attention than it got at the time. Because the tool learned from outcomes, you could point it at any outcome you cared about. A model for webinar attendance, another for churn risk, a third for upsell readiness. The same machinery, aimed wherever the business needed a prediction, while most teams were still arguing about whether an ebook download deserved 5 points or 15.Key takeaway: Pick one outcome beyond lead routing and build a scored view of it this quarter. Churn and upsell are the natural candidates because the outcome data already sits in your CRM. Pull 12 months of closed outcomes, find the 3 behaviors that most often preceded them, and hand that shortlist to the team that owns the motion.Why Lead Scoring Models Turn Into Black BoxesIf you have ever stared at the stars and flames in Marketo Sales Insight, or an intent dashboard announcing that an account is "surging," and wondered what any of it actually means, you have met the transparency problem. Darrell raised it directly on the show: as a user, he often had no idea what was going on inside these tools and no way to change it. Gary's answer starts with an admission. Breadcrumbs never fully solved this either.His best explanation for why the totems exist comes from riding Waymos in Austin. The first driverless ride freaks you out. By the third, you barely notice, and Gary credits the giant screen in the front of the Jag that renders the cars and pedestrians the vehicle sees.The flames work the same way. They translate a complex system into something a human can glance at and trust. The trouble is that really smart marketers refuse to stop at trust. They want to know what drives the representation, and then they want to override it. So every scoring vendor ends up refereeing between what the algorithm proved from historical data and what the marketer feels should count, and Gary says there is some validity on both sides of that fight.The model only knows where the business has been. A 10-year-old company that trains on all 10 years of data is muddying the waters, because who you sold to in year one barely resembles who you sell to now. And the moment you make a big deci...