SEMRON’s Five-Customer Strategy: Why Targeting Single-Digit Enterprises Beats Scaling to Thousands

SEMRON’s Five-Customer Strategy: Why Targeting Single-Digit Enterprises Beats Scaling to Thousands

Your board wants to see pipeline expansion. Your investors want compound growth metrics. Your VP of Sales is building out SDR teams and demand gen campaigns. And you’re sitting there wondering: what if we just need five customers?

In a recent episode of Category Visionaries, Aron Kirschen, CEO of SEMRON, an AI chip maker that’s raised $7.3 million, said something that makes most SaaS founders uncomfortable: “We do not or will not likely won’t have millions of customers. For us, it’s more like having five, six big entities that is generating millions of revenue.”

Five customers. Not five thousand. Not five hundred. Five.

This isn’t a limitation. It’s a strategy that completely rewrites how hardware companies should think about go-to-market.

The Math That Changes Everything

Traditional B2B wisdom says you need volume. More leads, more customers, more expansion revenue. The entire growth playbook—from PLG motions to account-based marketing—assumes you’re selling to hundreds or thousands of companies.

But when you’re building AI chips, the economics work differently. SEMRON’s technology serves a very specific need: running large language models on edge devices at a price point that makes sense. The companies that need this capability operate at massive scale. A single deployment could mean millions of devices.

This creates a paradox: your total addressable market measured in potential customers is tiny, but measured in revenue opportunity, it’s enormous.

When Aron says five or six big entities will generate millions in revenue, he’s not being aspirational. He’s being precise about who can actually deploy his technology at scale. These aren’t SMBs testing new tools. These are “very big, well known companies that are market leaders in their application or in that vertical.”

The implication for founders: before you hire your tenth sales rep, figure out your actual customer count. Not your TAM slide number. Your real, deployable customer universe.

What Sales Looks Like at Five Customers

Here’s where SEMRON’s approach gets interesting. Aron is explicit about what happens when your entire business depends on single-digit customers: “It’s very intense, very technical, very detailed, of course, but it is less standard marketing.”

Less standard marketing might be the understatement of the year. There’s no funnel to optimize. No conversion rates to improve. No A/B tests on email subject lines. The entire sales motion collapses into one core activity: “You have to sit down with their engineering team, with our engineers, and figure it out how we can make it work.”

This is engineering as sales. Your product team becomes your sales team. Your technical specifications become your pitch deck. Your ability to solve specific problems in their architecture becomes your competitive advantage.

The sales cycle doesn’t run through a CRM with automated touchpoints and nurture sequences. It runs through technical discussions where both sides bring deep expertise to solve a mutual problem. You’re not convincing anyone to buy. You’re collaborating to determine if the solution is viable.

For hardware founders, this changes hiring priorities completely. You don’t need a VP of Sales with enterprise SaaS experience. You need engineers who can sit in a room with a customer’s engineering team and earn their technical respect.

The Customer Engagement Model

SEMRON’s approach to customer engagement reveals what “intense” actually means in practice. They receive proprietary AI models from potential customers—the actual workloads these companies need to run. Then they deploy these models on their hardware, which they currently emulate with test structures.

“We try to make it fit into that,” Aron explains. “And so we discussing with them about, okay, what can we do about your models? And come back with a nice performance simulation.”

This isn’t a demo. It’s custom proof-of-concept work done before there’s a contract. It’s showing up with actual performance data on their actual models running on your actual hardware architecture.

The resource intensity is massive. Each customer engagement requires significant engineering time, custom simulation work, and deep technical collaboration. But when you’re targeting five customers instead of five thousand, this level of investment makes sense.

You can’t scale this motion. But you don’t need to. You need it to work perfectly five times.

When to Abandon the Traditional Playbook

Not every hardware company should adopt SEMRON’s approach. But certain signals indicate when a five-customer strategy makes more sense than traditional scaling:

Your product requires deep integration into customer systems. If deployment means months of engineering work, your customer count naturally stays low.

Your average contract value exceeds seven figures. At this price point, traditional sales funnels break down anyway. You’re not closing deals through demos and trials.

Your customers are deploying at massive scale. A single customer relationship could mean hundreds of thousands or millions of units deployed.

Your technology serves a specific, technical use case. When your product solves a narrow but critical problem for a small number of massive companies, volume plays don’t work.

The implementation requires ongoing technical partnership. If success means continuous collaboration between engineering teams, you can’t support hundreds of customers.

For SEMRON, all these signals align. Their 3D AI chips solve a specific problem—running AI models on edge devices at viable cost—for companies that operate at global scale. The physics of their business determines their customer strategy.

The Freedom of Constraint

What’s counterintuitive about SEMRON’s approach is how liberating it becomes. When you know your entire business depends on five relationships, you stop worrying about scalable acquisition channels.

You don’t need a content marketing engine generating thousands of MQLs. You don’t need a sales team cold-calling prospects. You don’t need to optimize your website conversion rate.

You need to identify the right five companies and become indispensable to them. Everything else is distraction.

This clarity reshapes resource allocation. Marketing budget doesn’t go to demand gen campaigns. It goes to deep technical content that reaches the specific engineers at the specific companies who make architecture decisions. Sales budget doesn’t scale with headcount. It scales with the depth of technical expertise you can bring to customer engagements.

Even hiring becomes simpler. You’re not building a sales organization. You’re building an engineering team that can earn trust and solve problems alongside customer engineering teams.

The Risk Profile

Of course, this strategy carries concentrated risk. Lose one of five customers and you’ve lost 20% of your business. Traditional wisdom says diversification protects you.

But Aron’s betting on a different calculation: the risk of spreading too thin across hundreds of customers who don’t really need your product exceeds the risk of going deep with a few who can’t succeed without it.

When SEMRON talks about enabling “live translation” or “personal assistant” capabilities that “enhance human productivity by factor of I don’t know, but not like 20%”—they’re describing capability shifts that major device manufacturers absolutely need. The risk isn’t that these five companies won’t want the technology. The risk is failing to deliver it.

For hardware founders evaluating this approach, the question isn’t whether you want five customers or five thousand. The question is whether your technology, your economics, and your deployment model naturally constrain you to single-digit enterprise relationships.

If they do, stop fighting it. Build the business model that fits the physics of your product. Sometimes the best scaling strategy is not to scale at all.