Founder-Led Sales: He Learned to Sell and Closed 50 Customers

Overview

Groundcover co-founder and CEO Shahar Azulay explains how the observability company grew to approximately $20 million in ARR, 250 customers, and 150 employees. The company competes with established platforms such as Datadog and New Relic by combining eBPF-based data collection with a bring-your-own-cloud architecture.

The conversation traces Groundcover’s evolution from an experimental product without its own user interface to a mission-critical replacement for incumbent observability platforms. Its first customer negotiated an initial $100,000 proposal down to $10,000, but the deal gave the founders experience, credibility, and valuable product feedback.

A central theme is founder-led sales. Azulay personally prospected through his network and LinkedIn, ran demos, supported proofs of concept, and negotiated contracts. He argues that early founders should prioritize closing customers and learning the complete sales process rather than optimizing contract value too soon.

The episode also examines the difficulty of turning an intuitive founder-led motion into a scalable sales organization. Groundcover had to learn how to hire and train account executives, sales engineers, and SDRs while developing the confidence, product maturity, migration process, and positioning required to replace Datadog outright.

Segmented Summary

[01:43] Groundcover’s product, market, and scale

[Fact] Groundcover provides a full-stack observability platform for R&D teams, helping them monitor the reliability of cloud applications and infrastructure.

[Fact] At the time of the interview, the company had approximately $20 million in ARR, 250 customers, and 150 employees.

[Fact] Product and R&D were based in Tel Aviv, while the go-to-market organization was primarily in the United States, with North America as its main commercial focus.

[02:30] Rethinking usage-based observability

[Fact] Azulay describes observability as a mature, mission-critical market in which most engineering teams already use an established solution.

[Fact] Cloud-native infrastructure, Kubernetes, and AI applications are increasing the volume and granularity of telemetry that organizations need.

[Fact] He argues that volume-based pricing changes customer behavior by encouraging teams to sample data, disable coverage, or leave parts of production insufficiently monitored.

[Fact] Groundcover instead charges according to infrastructure size, such as the number of monitored hosts, rather than the volume of logs, metrics, and traces ingested.

[Inference] The architecture attempts to separate the value of observability software from the rapidly growing cost of moving and storing telemetry.

[04:50] Bring-your-own-cloud and eBPF differentiation

[Fact] Groundcover manages its backend inside the customer’s environment, keeping the data plane on the customer’s premises while providing a managed experience.

[Fact] Because Groundcover does not host and mark up the telemetry, customers can retain more detailed data without paying the vendor according to ingestion volume.

[Fact] Its eBPF sensor gathers high-fidelity telemetry from the operating-system kernel with little or no application-code instrumentation.

[Fact] A monitored host receives logs, metrics, and traces as part of the offering.

[06:40] The founders’ experience with the underlying problem

[Fact] Azulay and co-founder Chez had worked as engineering managers and repeatedly encountered the same cycle: teams instrumented applications, then reduced or disabled telemetry because of cost.

[Fact] They did not view established observability products as technically poor; they believed the way customers consumed and paid for them no longer made sense.

[Fact] Their cybersecurity and deep-technology backgrounds led them to pursue a technical solution rather than merely offering the same service at a different margin.

[08:42] Testing the founding hypothesis

[Fact] The company began near the end of 2021, and the founders initially built a home lab to experiment with eBPF.

[Fact] After a few weeks of experimentation, they concluded that eBPF could provide unusually deep insight into applications and infrastructure.

[Fact] Bring-your-own-cloud was not initially framed as the company’s principal differentiator; it emerged naturally because the founders needed an architecture unlike Datadog’s.

[Fact] The first customer used a self-hosted infrastructure created by Groundcover roughly three months after the company started.

[10:34] Why eBPF became the core technical bet

[Fact] Azulay describes eBPF as an evolution of BPF technology that can safely run sandboxed logic inside the kernel.

[Fact] For observability, it can monitor network traffic, system calls, and other kernel events without embedding an SDK in each application.

[Fact] He compares this capability to an X-ray of software, particularly useful in cloud-native environments where teams use many programming languages and tools.

[Fact] Groundcover’s cybersecurity experience gave its early team relevant skills for building a secure, resource-efficient, high-scale host agent.

[Fact] Azulay says established observability vendors could not reproduce this capability overnight because developing an eBPF agent requires different expertise from building application SDKs.

[16:08] Winning the first customer without a finished product

[Fact] Groundcover’s founders used their Tel Aviv network to meet heads of DevOps and were introduced to the DevOps leader at Lemonade.

[Fact] The product did not yet have its own user interface; it consisted of the sensor, its telemetry, and the ability to create dashboards using open-source Grafana.

[Fact] Lemonade already used Datadog and initially saw Groundcover as a complementary tool for specialized cases, including DNS traffic and unusual cloud issues visible through eBPF.

[Fact] The DevOps leader who became Groundcover’s first buyer later joined Groundcover.

[17:34] The $100,000 proposal that became a $10,000 contract

[Fact] Azulay initially proposed a price of $100,000, partly because the founders knew an organization of that size could spend six figures on Datadog.

[Fact] The customer’s CFO offered a three-year agreement at $10,000 per year, and the founders accepted because they had no other customers or established negotiation process.

[Fact] Groundcover subsequently introduced public per-host pricing and discounted it heavily for early customers.

[Fact] Azulay believes founders should close their first dozen opportunities at almost any price because those deals teach them about procurement, security reviews, legal processes, quotas, and customer value.

[Inference] His argument is that early contracts should be treated as investments in organizational learning rather than reliable evidence of the product’s eventual market price.

[24:23] Founder-led sales to the first million in ARR

[Fact] Groundcover needed roughly 50 customers to reach its first $1 million in ARR and hired its VP of Sales after reaching several hundred thousand dollars.

[Fact] Azulay initially handled prospecting, discovery, demos, proofs of concept, pricing, and account-executive responsibilities himself.

[Fact] Even after the sales team began forming, he continued conducting demos for another year and a half to two years and remained involved in sales outside North America much longer.

[Fact] He says this kept him close to customers, competitors, technical use cases, and the company’s product direction.

[28:07] Why building the first sales team was difficult

[Fact] Groundcover needed time to define how to recruit, train, and evaluate technical sales engineers and account executives.

[Fact] The company’s first attempt to place a remote North American SDR under marketing, without a broader local team or established momentum, failed.

[Fact] Azulay says Groundcover made repeated mistakes while hiring AEs, sales engineers, and SDRs, although some early hires ultimately became senior leaders.

[Inference] The experience suggests that individual sales hires cannot substitute for a repeatable process, suitable training, and a functioning commercial center of gravity.

[30:24] How the first 50 to 100 customers were found

[Fact] Azulay’s early days usually included two or three prospective-customer demos, management of active proofs of concept, and extensive recruiting.

[Fact] He spent an estimated 30% to 40% of his time generating meetings, primarily by targeting heads of DevOps through LinkedIn, email, personal contacts, and conference relationships.

[Fact] On a productive day, he contacted approximately 10 to 15 people with personalized messages rather than relying on very high outbound volume.

[Fact] Network outreach, LinkedIn prospecting, Product Hunt, and follow-up with people who interacted with the product generated much of the initial customer pipeline.

[Fact] Early conversations were sometimes positioned partly as feedback sessions because prospective buyers did not yet perceive the young company’s outreach as a conventional sales call.

[35:40] Learning sales through repetition and mistakes

[Fact] Azulay entered the company as a technical founder without prior sales experience and found that enthusiasm for the product made early selling more comfortable.

[Fact] He learned through practice how to define sales metrics, set quotas, separate discovery from demos, and divide a founder’s combined responsibilities among specialized roles.

[Fact] He says there were no shortcuts: the company improved through experimentation, mistakes, reading, consultation, and conversations with experienced colleagues.

[Inference] Founder-led selling worked partly because the founder could modify the product roadmap and mobilize engineering resources in ways a newly hired salesperson could not.

[37:31] Becoming a full Datadog replacement

[Fact] Groundcover’s first 20 to 30 customers generally viewed it as either a greenfield platform or a complementary solution for use cases not covered by an incumbent.

[Fact] As bring-your-own-cloud became more prominent and the product approached feature parity, the company began positioning itself as a full replacement for Datadog and New Relic.

[Fact] By late 2024 or early 2025, its messaging, packaging, and product maturity had improved enough to support more frequent competitive evaluations and displacement deals.

[Fact] Average contract value increased as Groundcover moved from a supplementary, non-critical product to a mission-critical replacement for expensive incumbent platforms.

[Fact] By early 2026, the company had developed pricing strategies that allowed customers to begin switching several months before an incumbent contract renewal.

[42:06] Migration is part of the sale

[Fact] Some Tel Aviv customers began replacing Datadog with Groundcover in 2023, forcing the company to develop a formal migration and post-sales process.

[Fact] Groundcover initially failed in cases where customers bought its product but continued using Datadog for particular use cases.

[Fact] The company learned that closing a contract was insufficient; it also had to align expectations, support migration, and verify that the incumbent product was actually removed.

[Fact] These experiences influenced account-executive compensation and the company’s approach to post-sales execution.

[Fact] Azulay says it can be more productive to position the product correctly with new prospects than to persuade older customers to abandon the limited perception formed when the product was less mature.

[46:29] Confidence, pricing power, and knowing when to move on

[Fact] During 2025, Groundcover increasingly sold closer to list price and changed how it negotiated.

[Fact] Azulay says startups must recognize when they lack leverage, close an imperfect deal efficiently, and use the resulting customer, reference, and credibility to strengthen the next negotiation.

[Fact] He advises teams to keep reframing the product for future customers instead of dwelling on earlier deals that now appear underpriced or unsuccessful.

[48:19] Leadership lessons and rapid-fire recommendations

[Fact] Azulay’s principal leadership lesson is that a CEO should go as deeply as necessary into any function that is important, even performing work normally associated with an SDR.

[Fact] He recommends The Hard Thing About Hard Things, particularly for its account of the painful and lonely aspects of leadership.

[Fact] He considers distinctive field marketing, events, strong branding, and investment in US offices and workplace culture among Groundcover’s best expenditures.

[Fact] He minimizes recurring meetings and relies heavily on Slack and huddles for communication and task coordination.

[Fact] Outside work, cooking is a major passion; before returning to technology, he spent approximately a year and a half working in restaurants.

Podcast Commentary/Summary

The episode is especially valuable for technical founders learning to sell. Rather than presenting founder-led sales as charisma or a polished playbook, Azulay describes it as repeated contact with customers, direct ownership of every sales stage, and a willingness to close imperfect early contracts for the learning they produce.

Its strongest insight is the connection between technical architecture and go-to-market strategy. Groundcover’s eBPF sensor and bring-your-own-cloud design affect not only data collection but also pricing, product positioning, migration, and the company’s ability to challenge entrenched vendors.

[Inference] The account is less useful as a detailed operational guide to sales metrics, hiring criteria, or migration workflows because the interview discusses those areas mainly through lessons and anecdotes rather than specific frameworks.

[Inference] It is best suited to B2B SaaS founders, technical CEOs, and go-to-market leaders selling complex infrastructure products—particularly those entering mature markets where customers already use a mission-critical incumbent.