The Front Lines

Ofer Klein co-founded Reco, an AI SaaS security company that helps enterprises secure the AI agents and apps spreading across their SaaS stack. Along the way, Reco made two bets that could have gone badly. The first: stop being general purpose and double down on securing the SaaS layer, Salesforce, Workday, ServiceNow. One customer pushed back hard: "We built our practice on top of Reco. If you change, what should we do?" Ofer's team worked through it with them directly. 95% of customers from that period stayed, then doubled and tripled down. The second bet: go all-in on AI security, before most of the market believed in it. → Reco called itself "rico.ai" two years before ChatGPT existed → Their co-founder holds a PhD in AI and a PhD in cyber → The industry consensus at the time was to block AI entirely Ofer's team bet the opposite: AI's value was too big to stay blocked. They were right. Here's how they compress the sales cycle: tie the pitch to a mandate the CEO already funded. One customer had already committed to save $1 billion using AI internally. Once Ofer connected security to that goal, it became a board-level conversation. Another pattern: strict AI bans don't lower risk, they just hide it. Across customers, roughly 80% of AI usage was happening in the shadows, outside any sanctioned tool. As Reco moved from commercial deals (2,500 employees and under) to enterprise accounts (25,000+), they had to replace sellers and managers who'd excelled at the smaller stage but weren't built for the next one. Ofer traces some of this discipline back to flying in the Israeli Air Force: brief, execute, debrief, adjust, now applied across the whole company, not just himself. Listen to the full conversation with Ofer Klein on BUILDERS to hear how Reco built its go-to-market playbook for AI security: https://lnkd.in/eivu8w7T

Betsy McKibbin's sales team gets a Monday morning wake-up call from an AI version of her. Here's the story: Betsy started as a software engineer. Computer science undergrad, she loved building and shipping code. Then she found something she loved more: understanding why people buy. That took her into marketing. Almost 20 years of it now, across HR tech and edtech, before she landed in AI. She joined Yoodli AI Roleplays about a year ago. Employee number 32. Today, the team is over 80. Yoodli started as an AI roleplay tool for sales reps to rehearse conversations. It's grown into an always-on AI coach used across sales, retail, healthcare, legal, and finance. Doctors use it to practice delivering hard news to patients before the real moment. Retail teams use it before a product launch hits the floor. Google Cloud used it to train 15,000 reps in a single month. But Betsy's biggest problem wasn't external. It was internal. Her own sales team couldn't keep up with how fast Yoodli itself was shipping. New features constantly. Reps getting asked about updates they hadn't even heard of yet. The old playbook: write a product brief, present it to the team, hope it sticks. Not fast enough anymore. So Betsy and her head of product built something new: an AI agent that scans the team's Asana, Slack, and GitHub for what just shipped, then teaches it to the sales team automatically and lets them practice talking about it immediately, in the same session. They call it Betsy AI. It wakes reps up on Monday mornings: here's everything that shipped, let's practice what you need to say this week. Now she wants to extend that same loop to partners, and eventually to customers directly. On growth, Betsy's team doesn't only lean on paid channels. One story: a national broadcast crew set up two blocks from Yoodli's office during the World Cup in Seattle. A marketer who'd been on the job two weeks grabbed a cardboard sign and got the company name into the shot. Site visitors peaked that day. Her advice to other marketing leaders, especially in AI: as everyone on your team starts generating their own content with AI tools, you have to draw a hard line on what marketing owns, or you lose your brand voice entirely. Full conversation is live now on Marketers of Technology https://lnkd.in/dmYYWQpg

RapidScale's CMO, Jessica Maria has a rule: if AI finds your point of view, you've already lost. Here's how she is rebuilding a brand around that idea: Jessica never planned to end up in marketing. She wanted to write for magazines and publish books. Then she started a family, needed a different cadence, and fell into corporate comms work instead. Writing for companies like Microsoft and Adobe paid better than writing magazine articles. That progression eventually landed her as CMO of RapidScale, a managed cloud services and professional services provider working across public, private, and hybrid clouds. When she stepped into the role, RapidScale's brand was built around one idea. We'll help you navigate the chaos of technology. But she noticed the business had evolved faster than the story around it. So her team ran a real test. In 2025, RapidScale brought in Cunningham Collective to survey clients on how they actually perceive the company. The result came back unprompted and consistent. Every client called RapidScale their partner. Not their vendor. Not their supplier. That single finding became the anchor for RapidScale's entire brand shift. Now Jessica has a strict rule for where AI fits into that story. If a company uses AI to discover its point of view, she says, it's already lost. Prospects can run that same search themselves in five seconds. Where AI earns its place: researching a buyer segment's pain points, market intelligence, competitive intelligence. Where it doesn't: writing the point of view itself. Her head of solutions marketing has a name for the opposite mistake. Trying to "boil the ocean" in every piece of content. Her fix: find the one core point everything else sits on top of. Build outward from there. RapidScale's marketing org has grown to close to 20 people, split across two pillars: brand and demand, with solutions marketing (including field marketing) layered on top. Next up: the company is relaunching its website, moving off a .net domain onto rapidscale.com for the first time, targeting the first week of July. Her 2026 bet: hyperpersonalization on the demand side, paired with continued investment in thought leadership and brand. Her logic: tools and platforms get commoditized. A genuine, anchored point of view doesn't. New episode of Marketers of Technology, out now. Listen here: https://lnkd.in/e_2_jKnh

GlobalComix is trying to digitize the last major entertainment category that never fully went online: comics. Henrik Rydberg took over as CEO of a platform that pulls together content from 330+ publishers and 25,000+ independent creators, all reading in one place. Here's how they got publishers to actually adopt it: The hardest sell wasn't the reading experience. It was the licensing. → Every publisher works with different partners → Across different countries → With different rights granted to each one So instead of leading with the product, GlobalComix leads with the strategic problem. Pitching the licensing team directly? "The most boring way for us to enter," and where they see the least success. The real traction comes from bigger conversations, like helping Japanese manga publishers fight piracy. Less than 1% of manga has ever been officially translated into English. Publishers who tried building their own platform ran into the same wall: → It was easy to build the software → But they didn't have enough content to make it compelling → Catalog depth, not tech, turned out to be the real barrier The publishers most willing to go digital weren't the cautious ones. They were the ones confident enough in their IP to see online exposure as growth, not cannibalization. A reader who discovers a story online still buys the hoodie. Still buys the movie ticket. That's the bet. GlobalComix doesn't measure itself against other comics platforms. It measures itself against every other place a reader could spend their attention, the same logic behind Netflix competing against sleep. Listen to the full episode with Henrik Rydberg, CEO of GlobalComix, on BUILDERS to hear how they're building the infrastructure behind digital comics: https://lnkd.in/egxJ4iCG

Arize AI's open source AI evaluation tool, Phoenix, has been downloaded millions of times. Its VP of Marketing spent the first years of her career designing Google's homepage logo. Here's how she got from art history to AI observability, and the GTM playbook she's running today: Leah Clark studied art history in college, planning to work in the art world. Her first job was in traditional newspaper advertising, before she found her way onto Google's sales team. Within a few years, she moved into marketing there, and eventually onto the team that designs the Google Doodle, roughly 300 of them a year. From there: product marketing at Twitter, then brand marketing at Medium. Then GitHub, for almost 8 years, as their first brand marketing hire. She built out social, content, swag (internally called "Shop"), and enterprise marketing motions. After 8 years at one company, she wanted to see the whole marketing function, not just one piece of it. So she joined Arize, an AI observability and evals platform, to build and run marketing from the ground up. Arize's business spans two very different audiences: → Phoenix, their open source product, with millions of downloads → A consumption-priced SaaS motion running from small teams to Fortune 500 accounts Her approach to standing out in a crowded AI category: stop copying competitors' campaigns and events. Instead, root the brand in what's structurally true about the company. In Arize's case, two technical co-founders obsessed with education and bringing non-experts along. She's also betting on channels a lot of SaaS marketers have written off. Arize ran billboard campaigns in San Francisco and New York, and measured the lift the same way you'd measure a paid campaign: comparing homepage traffic, signups, and search volume in-market before and after launch. Her framing: people have gotten very good at scrolling past digital ads. What can't be scrolled past is a billboard, a swag drop, or a pop-up space you can actually walk into. The line she keeps coming back to, from her time at Google: know the user, know the product magic, connect the two. Listen to the full episode of Marketers of Technology here https://lnkd.in/eNHEEcxm

Abhishek Agarwal left a stable Head of Sales role because he wanted to be at the center of the AI Revolution. Nine months later, he's Global CRO & EVP at Successive Digital, the parent company behind Kagen, an AI-native platform engineering company running in 10 countries with 150-175 customers, 15 of them already live on the AI platform. The pivot wasn't just external. Before Successive Digital sold AI transformation to customers, it had to run through one internally: → Shift the org from delivering traditional technology services to building AI products and platforms, for both the company and its customers → Compress build cycles from 12 months to weeks by pairing agentic development (AI does the building) with humans in the loop for requirements and validation, and humans on the loop for release → Bundle AI FinOps into every rollout after watching customers "blow out" AI budgets on unmanaged token consumption, the same mistake companies made with cloud costs a decade earlier → Rebuild the hiring bar around adaptability instead of AI tenure, since demanding a decade of experience in something that's existed for about a year filters out exactly the people who'd be good at it Three decisions shaped how this actually worked: 1. Build it before you sell it. Successive Digital uses AI to build, market, and sell before pitching that same transformation to customers. That's what makes the pitch credible instead of theoretical. 2. Evangelism can't sit with one function. Customers don't want AI validation only from a CRO. They want to hear it from the CFO, the COO, the director of delivery, so the internal rollout deliberately built multifunctional teams instead of leaving AI to the tech function alone. 3. Data debt doesn't get "fixed" before you move. Rather than waiting for clean data, the approach is human-plus-AI working against a semi-cleansed view. Waiting for perfect data before adopting AI just costs competitive advantage. Listen to the full episode with Abhishek Agarwal on The Sales Front Lines: https://lnkd.in/eyCY3GEa

OpenTAS turned one of the most boring categories in B2B tech into a story people actually stopped for. Here's how: OpenTAS builds business process software for the oil and gas supply chain, specifically the segment few marketers ever touch: terminal management. There are 15,000 tank farms worldwide. Only 4,400 of them are addressable with modern software. OpenTAS already serves 86 of them, built on Oracle technology. They set out to build a new terminal management platform on the Microsoft stack, hoping to repeat what they'd already pulled off in the downstream segment: getting their modules onto SAP's price list and letting SAP's own account executives sell them straight into BP, ExxonMobil, and other majors. A new platform needed a new story. So CMO Wolfram Wege brought in Brandigans, an outside creative network led by Grant Price, to answer one question: how do you make terminal management interesting? Their first move: drop the word "solution." Every competitor already used it. They repositioned the product as a "platform" instead, built to replace the entire stack of legacy tools operators had bolted together over decades, not add one more tool to the pile. The slogan: "Complexity removed." At one point, OpenTAS's own product team tried to soften it to "complexity managed." Wolfram refused. The softer version lost the point entirely. Before showing the board the actual rebrand, he opened the pitch with Apple's 1984 ad, the one where a woman smashes a screen with a hammer, to set the tone before the reveal: go bold, or don't go at all. The board said yes. The campaign was built around one visual contrast: the old systems shown as a disconnected high-rise, the new platform as a single gold surface everything sits on. They also ran a print ad that broke the industry's own visual habits, a single flower growing through cracked concrete, inside a trade magazine otherwise filled with gray industrial imagery. They presented the rebrand at two of the industry's largest terminal expos, one in Rotterdam and ILTA in the US, and the response, in their words, was amazing. Two years later, the rebrand is still holding. Wolfram and Grant have worked together since 1998, long enough to hand off a half-written script and know exactly what the other means by it. Their real lesson: nobody buys a system. They buy a vision they can picture working before they've seen a single feature. Full episode of Marketers of Technology is out now: https://lnkd.in/eMfpE7MW

Sarah Monette walked into Cosine three and a half months ago with zero ML background. No technical co-founder whispering strategy in her ear. No playbook for marketing an AI coding agent to defense and healthcare buyers. Just a generalist marketer, a 30-person team, and a category dominated by OpenAI and Anthropic. Here's what she did first. She launched a campaign called "Built for Engineers with Taste." The messaging was deliberately divisive. It was explicit about who the product wasn't for — including herself. It landed exactly because of that. Her logic: if you're trying to please everyone, you're going to please no one. So she stopped trying. Then came the GTM motion. No outbound. No content spray. The primary channel was partnerships — building one or two deep relationships with consultancies that already held the trust of her target accounts. Those partners assembled the dinners. 10 to 15 people. In person. In 2026, that's the move. Because emails are ignored. LinkedIn posts are AI-generated. Video can be faked. The one thing that still builds trust in a regulated enterprise sale is repeated face-to-face contact. Old motion. Highest ROI motion. Then there was the government angle. The UK launched a sovereign AI fund and selected six organizations to receive backing. Cosine was one of them. When you're trying to close defense and healthcare deals, a government seal of approval does more than any case study. And when the US restricted access to Fable 5, the entire conversation around AI sovereignty shifted overnight. Cosine was already positioned there. Meanwhile, her head of growth — a former developer — was running unbranded landing page experiments to test whether people would pay for new products before they were ever built. No Cosine branding. Just a value proposition and a price point. Let the data decide. Sarah also built what she calls a GTM brain. Every week she feeds it: news updates, product changes, customer signals. It tells her what campaigns to run and what narrative to lean into. A lean team running faster than most companies 3x their size. Three and a half months in, one of her engineers got stopped on a plane because someone recognized the Cosine backpack he was wearing. That's the goal. When people think OpenAI, Cursor, Anthropic — she wants Cosine to be the next name on that list. New episode of Marketers of Technology is live. Listen here: https://lnkd.in/eC5Vb4J5

Kodem's founders built the exploit that first hacked the iPhone. They can't talk about it publicly. Not on the website. Not in a press release. Not in a funding announcement. Here's how their CMO built a marketing engine around a story he's not allowed to tell: Mahesh Babu joined Kodem knowing the founders came from NSO Group, the team behind Pegasus. His first instinct was to put the founders front and center: a homepage built around the people who created Pegasus. The founders said no. Too risky. Back to the drawing board. So the story went underground. Nothing on the site. Nothing in press. The only place it gets told is in a closed room, at events Kodem calls "backroom briefings." Then came the real breakthrough, not a marketing insight, a technical one. On a call with one of the founders, Mahesh asked how Pegasus actually worked. The answer reshaped the entire product story: → The exploit used low and medium severity bugs, the ones vendors never bother fixing → One bug alone did nothing. Chaining bugs together is what let them move from one app to full device access → Attackers only care about what's reachable from the outside. A vulnerability nobody can reach isn't a real risk Kodem rebuilt its own product narrative around that same logic: don't just scan for vulnerabilities, show which ones are actually reachable and exploitable in production. Mahesh tested the message live, on stage, at a security event. He asked the room: how many of you have an iPhone? How many have heard of Pegasus? About 60% raised their hands. "I work for the creators of Pegasus." He tracked the message across four events in the same city. People remembered it every time. Meanwhile, roughly 18 other funded startups were all calling themselves the same thing: ASPM, the acronym Gartner coined for a market Gartner had split into 15 sub-categories. Kodem refused to use it. Instead of fighting to define a new category, Mahesh anchored Kodem to the category customers already understood and already had budget for: code scanning. Kodem still doesn't have a category label for what it sells. It doesn't have a booth at Black Hat either, and never will. Full conversation with Mahesh, CMO of Kodem, on The Marketing Front Lines here: https://lnkd.in/e5i_Vznu

Ardie Sameti spent a decade scaling Accolade from a handful of customers to over 20 million lives covered. Then he walked away to start over. Here's the story: At Accolade, Ardie ran AI and platform efforts as the company scaled through acquisitions to expand its reach. But every acquisition brought a different tech stack, a different data set, more fragmentation. → Operations became a cost center → Teams spent their time firefighting instead of building → The problem wasn't unique to Accolade Ardie noticed the same pattern at companies of every size, from Uber to Walmart. So he and Accolade's CEO and chairman, Rajeev Singh, made a decision: spin off and build the fix themselves. That became Scala. Rajeev as co-founder and executive chairman. Ardie as co-founder and CEO. The hardest part wasn't the business plan. It was the conversation with his wife, with three young kids at home. She didn't hesitate. She told him to do it. Inside Scala, the operating philosophy looked nothing like Accolade's playbook: → They were ahead of their own hiring plan, and chose not to hire anyway, to keep direct access to prospect feedback → Every closed deal gets traced back to its real source before they scale spend on that channel → AI avatars run outbound, but only against active buying signals, never blasted at scale Ardie's view on what actually creates an edge in 2026: frontier models are a commodity. Anyone can build an agent. The moat is the proprietary data and workflow context a company builds inside its customers' operations. Scala's bet: the workforce goes hybrid, agents and humans working side by side, with humans supervising fleets of AI doing the work that used to bury operations teams in busywork. Listen to the full conversation with Ardie Sameti on BUILDERS to hear how he's building Scala's go-to-market from scratch, the AI-native way: https://lnkd.in/eDPYruy3