Yuval
Shiboli

GeoEdge helps publishers detect and block malicious ads. We published quarterly threat reports, but anonymized the data to protect relationships with ad platforms. For publishers trying to evaluate their partners, that made the reports interesting but not actionable.

For our next report, I flipped the framing. Instead of a threat report, we published a performance benchmark: the top-performing ad networks, scored across three dimensions: user protection, content quality, and user experience. I built the scoring methodology on our own detection data, automated it in Amazon QuickSight, and established an industry benchmark threshold that determined who made the list. The result was a report with real methodology behind it, not just opinions.

Publishers got something immediately useful. Top-ranked networks started reposting the report publicly. Networks not on the list reached out proactively, opening new commercial conversations.

GeoEdge and AppHarbr serve five distinct customer types, each with different business models, priorities, and definitions of value. With one generic narrative across both products, the messaging was speaking to everyone and resonating with no one.

Built a full messaging framework for each segment: the specific pain they feel, the value proposition that maps to it, and the proof points that make it credible. Each value proposition is tied directly to product features, so the sales team can demonstrate exactly how the product delivers the value. Translated into segment-specific website pages and tailored sales decks.

The segmentation work gave the sales team a sharp, relevant narrative for every customer conversation, from the first touchpoint to the close.

In a market where every competitor claims the same thing, "we're better" is not a differentiator. Sales had no structured way to explain why GeoEdge was different, competitive knowledge lived in individual inboxes, and there was no shared understanding of how to win against specific competitors.

Conducted systematic research across 6-7 competitors: obtained product access, attended competitor demos, analyzed sales collateral, pricing, and public communications. Applied trackers to the company's call recording repository to flag every mention of a competitor name, turning the field's institutional knowledge into structured intelligence. That research exposed specific weaknesses in each competitor's approach and translated into battle cards, updated decks, and a competitive module in onboarding, with regular refreshers.

Sales could now walk into a competitive situation with a clear, specific narrative: not "we're better" but "here's what we do differently and why that matters for you." The team started winning more competitive deals and took several accounts directly from competitors.

When I joined as the first PMM, there was no structured way to get new salespeople up to speed, no shared place where content lived, and no reliable channel between the product team and the people selling and supporting it. Everyone was working from their own version of the story.

Built three connected systems: a role-specific onboarding and training program for every SDR, AE, CSM, and SE; a searchable internal library where all content lives, structured enough to answer specific questions on demand; and a formal product-to-field channel that briefs the team on new features before launch, framed around customer value, and routes feedback back to product.

New hires ramp faster. Customer-facing teams speak the same language. And when a customer's feature request ships, the CSM can close that loop directly.

GeoEdge's security team was detecting sophisticated, novel attacks that no one else in the industry was seeing. But researchers are focused on the technical work, not the marketing value of sharing it. The findings stayed internal, and the company's security credibility stayed invisible.

Brought the security team along on the idea of sharing their work externally, then built the infrastructure to make it repeatable: a data tracking system to surface patterns over time, and a process for translating deeply technical findings into stories non-experts could follow. When a researcher detected LANJack, a sophisticated attack using ads to probe local networks and IoT devices, we announced the discovery to coin the name and submitted the full research to Blackhat.

The Blackhat submission was accepted. GeoEdge will present at Blackhat US 2026, one of the most selective security conferences in the world.

CSMs had no systematic way to know which accounts needed attention and why. Account health was assessed in quarterly reviews based on gut feel, by which point issues had already unfolded or opportunities had already passed.

Built personalized Amazon QuickSight dashboards for each CSM, tracking real-time KPIs across every account: impressions monitored, incident and blocking rates, demand sources, and coverage. ML-based anomaly detection alerts CSMs the moment something behaves unexpectedly, before it becomes a problem the customer notices. The dashboards also flag changes to protection settings that might signal a growth conversation, and compare usage against billing packages to surface upgrade opportunities and churn risks.

CSMs shifted from reactive to proactive. When GeoEdge blocks an attack the customer never saw, the CSM can show them exactly what happened and why it matters. That changes the conversation from "are you getting value?" to "here's the value we delivered last night."