Between Two Quarters: How to Sell into Invisible Markets with Roshan Patel

Key takeaways from this episode:

• Hire your first sales rep when your outbound motion reliably produces meetings every week and you can document the process in a Google Doc.
• Early sales reps don't need industry experience; they need experience selling to the same customer size. Sales leaders need exact-market experience.
• Conferences work for many industries, but need to optimized for direct meetings (dinners, etc) not booths.
• AI made shallow personalization worthless and deep personalization affordable. Arrow's best outbound asset is a 12-page deck researched for a single prospect.
• In markets where buyers are unreachable digitally, GTM is a trust problem, not a funnel problem.

Every GTM playbook we've ever seen shares one assumption: you can reach your buyer.

But what if you can't? What if your buyer isn't on LinkedIn, their email is buried, and the person who actually runs the business has a limited digital footprint?

That's the market Roshan Patel chose in building for healthcare. But, there are many similar under-digitized vertical markets where this dynamic exists.

For this episode of Between Two Quarters, we flipped the script and sat down with a founder who sells for a living. I've known Roshan Patel for over three years as the CEO and co-founder of Arrow, an nvp capital portfolio company using foundation models to automate revenue cycle management for small to mid-size healthcare providers. It's one of those services industries that is an excellent target for AI, but only if you can also solve the data collection and cleaning challenge, and really understand the complex workflows around healthcare billing, insurance reimbursement, and collections.

Roshan came to the problem honestly: a former VC who grew up handling billing for his parents' medical practice, in a family where everyone went to med school except him. He is exceptionally good at identifying commercial opportunities in healthcare, and one of the things I've admired watching him and his co-founder Yash Joshi build is how directly they connect real-world opportunities to product.

Arrow sells to mid-market physician groups: three to fifty-plus locations of specialty practices. At the lower end these are small businesses run by people who trained for a decade in medicine and zero days in business, in an industry where you deliver the service on day one and get paid 60 to 90 days later.

What I really wanted to dig into with Roshan was the transition every founder dreads: going from "found a problem, built a product, got some customers" to building an actual sales team. Moving from founder-led to team-led sales is genuinely hard, and there are breaking points all around it. The first AE. Then a sales leader. Then the first hires under that leader.

Across everything Roshan shared, one pattern kept showing up:

When your buyer is invisible to digital channels, distribution advantage doesn't come from better tooling. It comes from physical presence, borrowed trust, and doing things that don't scale, on purpose.

Why Cold Email Stopped Working (and Why That's an Opportunity)

The math on outbound in 2026 is uncomfortable. AI made personalization at scale available to everyone, which means personalization at scale is now worth roughly nothing. Buyers have adjusted, and the filter is instant.

In most markets, that's an annoyance. In healthcare, it's closer to terminal, because the channel was barely open to begin with. Think about how hard it is to get your own doctor on the phone. Now try selling them something.

Roshan's take: the founders who win in these markets aren't the ones who out-tool the noise. They're the ones who exit the channel entirely.

Which is exactly what Arrow did.

When Should You Hire Your First Sales Rep? (The Google Doc Test)

For the first two or three months, Roshan and his co-founder did exactly one thing: walked door-to-door through New York City doctors' offices. They borrowed the idea from Zocdoc's early days, and at first it produced exactly what you'd expect: "please get out, you don't have an appointment."

Then they noticed who does get past the front desk: pharma reps. And pharma reps never show up empty-handed.

So the motion became: find a Dunkin' Donuts, buy a dozen, walk into a practice, repeat every ten minutes. Find the Manhattan office buildings where every floor is a healthcare provider and work the whole building until security has opinions.

This was the part of the conversation I'd been waiting for, because it gets at something founders consistently miss about "do things that don't scale." The unscalable phase isn't the strategy. It's how you discover the strategy. And Roshan's bar for when that phase is over is the cleanest we've heard on this show:

Hire your first sales rep when your outbound motion reliably produces meetings every week and you can write down how to do it in a Google Doc. If you can't document it, you haven't found a repeatable motion. You've found luck.

The doc is the product of the donuts.

Are Conferences Worth It for Early-Stage Startups?

Short answer from Arrow's experience: yes, but.

In underdigitized industries, connecting with buyers in person is critical. Conferences are that + they exist to bring together industry people to connect. That said, buying a ticket to attend or spending on a booth are too passive to create quality conversations. Arrow's first booth taught them what most founders learn expensively: at a financial healthcare conference, every booth looks identical, and nobody remembers any of them.

So they inverted it. Instead of buying a booth, they host a private steakhouse dinner during the conference: curated guest list, real conversation about the problems practices are facing, and prospects who are genuinely grateful to have something to do when the sessions end.

Then came the sharper realization. The conference had already done the expensive work of aggregating their exact buyers into one city on known dates. The badge was optional. Now Arrow targets a conference, shows up in the same city the same week, and hosts the dinner, once a month. It's their best-performing channel.

This rhymes with what Liz Ward told us about how enterprises actually buy from startups: in vertical markets, technology doesn't spread through cold outbound. It spreads through operator-to-operator credibility. Roshan sees the same physics in New York, where practices in the same specialty all know each other. One behavioral health customer, won through a cold DM in a founder Slack group, opened an entire vertical for Arrow.

Should Your First Sales Hires Have Industry Experience?

The first AE is the first of those breaking points, and there's no one-size-fits-all approach. But I love the way Roshan thinks about the process, because it exposes an asymmetry most founders get backwards.

None of Arrow's three sales reps came from healthcare. Roshan hired for hunger, grit, and comfort with rejection, plus one specific filter: early reps don't need industry experience, but they do need experience selling software to the same customer size. A rep who spent a career prospecting into ten enterprise accounts runs a fundamentally different motion than one working a universe of thousands of small practices. Industry language can be learned. Motion is muscle memory. I've seen this dynamic play out across a number of different companies in separate industries, with similar conclusions. Of course there is the hope that you'll find an AE with some ready to go pipeline from a prior role, but that's rare.

But when Arrow hired its head of sales this year, the logic flipped completely. There, healthcare experience (selling the same product category to the same buyer) was non-negotiable. The playbook already worked. Roshan didn't want someone who'd take six months to figure it out. He wanted fuel on the fire.

Two footnotes worth stealing:

• By Roshan's count, roughly a dozen people on earth fit his head-of-sales profile, and none of them answered his LinkedIn messages. The same people responded immediately to a recruiter. For scarce senior talent, the channel is the credibility.
• His head of sales still sells: half managing, half running demos and discovery alongside Roshan. At this stage, a leader who only manages is managing too little pipeline to matter.

The Sale Happens in Discovery, Not the Demo

Roshan had never sold anything before Arrow, and his first calls followed the default founder script: pitch, demo, ask, lose.

The upgrade, absorbed from an obsessive diet of sales podcasts like 30 Minutes to President's Club, was refusing to skip discovery, even when the buyer resists it. And healthcare buyers always resist it. The doctor gets on the call and says: I have ten minutes, what have you got for me?

The fix is procedural, not clever. Set the agenda in the first thirty seconds: you'll absolutely see the demo, but first, a couple of questions about the business. Keep them open-ended. Small talk before you ask about revenue.

Because once a buyer has said their pain out loud (how big it is, what it's costing them, why it matters now), the demo stops being a pitch and becomes the answer to a problem they just described. In Roshan's words, the sale almost does itself.

And in Arrow's market, discovery almost never comes up empty. Once a practice takes the meeting, it's near-certain they have a pain point around getting paid, or getting paid faster.

The hard part was never the pain. It was the meeting.

AI Didn't Replace the Old Playbook. It Armed It.

AI destroyed shallow personalization and made deep personalization affordable, at the same time.

Arrow's current outbound unlock is a 12-page deck built for a single prospect. Real numbers assembled from public data about that specific practice. The practice's name on the title page. Addressed to the actual decision-maker. Sent cold.

The reaction it produces is the entire point: this isn't spam. Someone actually studied my business.

That's not a different strategy from the donuts and the dinners. It's the same one, proof of genuine, individual effort in a channel flooded with fake effort, executed with better tools.

The Real Insight

In markets where the buyer is invisible, GTM isn't a funnel problem. It's a trust problem.

And trust, it turns out, has a fairly consistent price:

• Show up physically, where your buyer already is
• Bring something: donuts, dinner, or twelve pages of homework
• Let operators sell to operators
• Do it unscalably until you can write it down. Then, and only then, scale it

The founders who treat "unscalable" as a phase to survive get a playbook out the other side. The ones who skip it get a booth in a sea of booths.

Check out the full conversation below, and for more of nvp capital's Between Two Quarters, visit https://www.nvpcap.com/content