The GTM Playbook We Built With Zero SDRs

We have zero salespeople.

No SDRs. No AEs. No sales engineers. No revenue team Slack channel with motivational GIFs. At Well, twelve people build the product, and the product builds the pipeline.

This is not a humble brag. For the first six months, it terrified me.

Every founder I talked to was hiring their first SDR by month three. Investors kept asking when we'd "build out the go-to-market team." The playbook was obvious: hire two SDRs, give them a dialer and a Salesforce instance, point them at a list. That's how you show traction. That's how you fill a pipeline. That's how serious companies operate.

We did none of it. Instead, we built a machine.

The Thesis: PLG Is the Only Viable GTM for AI-Native Startups

Here's the argument, stated plainly: if your product requires a human to explain its value, you haven't built a good enough product.

This is especially true for AI-native companies. The whole premise of AI is that it reduces the need for human intervention. If you need a sales team to convince people to use your AI tool, you've already contradicted your own thesis.

Product-led growth forces a constraint that makes your product better. When no one is there to walk a prospect through onboarding, your onboarding has to be flawless. When no SDR is following up to "check in," your product has to deliver value fast enough that users come back on their own.

PLG conversion rates are brutal. Industry benchmarks hover around 2-5% for free-to-paid. That number should scare you. It scared us. But it also clarifies your priorities: every engineering hour goes into making the product stickier, not into building slide decks for a sales team.

The math is simple. An SDR costs $80-120K fully loaded. Two of them, plus a manager, plus tooling, and you're spending $350K+ annually before they've booked a single meeting. For a twelve-person startup burning through a seed round, that's not a go-to-market strategy. That's premature scaling disguised as ambition.

The Pipeline We Actually Built

Zero salespeople doesn't mean zero outbound. It means automated outbound. Here's exactly how our pipeline works, tool by tool, so you can replicate it.

Layer 1: Intent Signals — Theirstack

Theirstack monitors technology adoption signals across companies. We track when companies add or remove specific tools from their stack, tools that indicate they're solving the exact problem Well addresses. This gives us a warm list of companies actively in-market, not a cold list scraped from LinkedIn.

We filter for company size (20-500 employees), geography (North America and Western Europe), and specific technology signals that correlate with our ICP. This produces roughly 200-300 new companies per month.

Layer 2: Website Visitors — Snitcher

Snitcher identifies the companies visiting our website. Not individual people, companies. When someone from Acme Corp reads our docs or pricing page, we know Acme Corp is evaluating us.

This is the highest-intent signal in our pipeline. Someone already found us, read about us, and self-selected. Around 15-20% of our Snitcher-identified companies overlap with our Theirstack signals. When we see both signals, that account goes to the top of the queue.

Layer 3: LinkedIn Engagement — Reactin

Reactin tracks who engages with our LinkedIn content: likes, comments, shares. These people have raised their hand. They've seen our thinking, found it relevant, and taken a micro-action. It's not a buying signal, but it's a relationship signal.

We use Reactin to identify individuals within target accounts who are already warm to our ideas. This turns cold outreach into a warm conversation: "Hey, I noticed you commented on our post about X. We're building exactly that."

Layer 4: Orchestration — Cargo

Cargo is where it comes together. It's our orchestration layer, the glue that connects Theirstack, Snitcher, and Reactin into a single automated workflow.

Here's what Cargo does:

1. Ingests signals from all three sources daily
2. Enriches contacts with firmographic and contact data
3. Scores and prioritizes accounts based on signal overlap
4. Triggers personalized email sequences (not templates, sequences built from the specific signals we observed)
5. Routes high-priority accounts for manual review The enrichment step is critical. Cargo pulls in job titles, company funding stage, team size, and technology stack. Everything we need to personalize outreach without a human researching each account.

Layer 5: CRM Pipeline — Attio

Everything flows into Attio, our CRM. Attio is where signals become pipeline stages. An account might enter as "Signal Detected," move to "Engaged" when they reply to an email, and progress to "Design Partner Candidate" when they book a call.

We chose Attio over Salesforce or HubSpot because it's designed for the way modern teams actually work: flexible data models, no bloated enterprise features we'd never use, and an API that doesn't make you want to quit the industry.

The Design Partner Program

The pipeline doesn't feed a sales process. It feeds a Design Partner Program.

Instead of trying to close deals, we invite high-signal companies to co-build the product with us. Design partners get early access, direct influence on the roadmap, and a monthly email update from me personally. Not a marketing newsletter, an actual letter explaining what we shipped, what we learned, and what's coming next.

This reframes the entire relationship. We're not selling to them. We're building with them. The conversion from design partner to paying customer isn't 2-5%. It's closer to 40%, because by the time we launch a feature, the people who asked for it are already using it.

Currently, we have 34 active design partners. Eighteen came through the automated pipeline. Nine found us organically. Seven came through referrals from existing partners. The pipeline is the primary acquisition channel, and no human touched it until the first reply.

The Numbers

I promised real numbers, so here they are. These are from the last full quarter:

• Theirstack signals processed: ~850 companies
• Snitcher-identified visitors: ~1,200 companies
• Reactin engagements tracked: ~2,400 individuals
• Cargo-enriched and scored accounts: ~620
• Automated email sequences triggered: ~480
• Replies received: 67 (14% reply rate)
• Meetings booked: 28
• New design partners added: 11
• Total cost of pipeline tooling: ~$1,800/month Compare that $1,800/month to the $30K+/month you'd spend on two SDRs. The pipeline runs 24/7, doesn't take PTO, doesn't need a manager, and improves every month as we tune the signals.

The reply rate matters. Fourteen percent is high for outbound. Industry average for cold email sits around 1-3%. The reason ours is higher: we're only reaching out to people who have already shown intent. Every signal layer filters the noise, so by the time someone gets an email from us, it's relevant.

What Scared Me About This Approach

I'll be honest about the fear, because I think it's the reason most founders hire SDRs too early.

The fear is legibility. When you hire two SDRs, everyone, your investors, your board, your team, can see that you're "doing sales." There's activity. There are calls logged, emails sent, meetings booked. It looks like progress. It feels like a real company.

An automated pipeline is invisible. Nobody claps when Cargo enriches 200 accounts at 3 AM. There's no weekly sales standup to make everyone feel like the business is moving. The work happens in the background, and the results show up in the CRM without fanfare.

For six months, I kept second-guessing the approach. Were we leaving pipeline on the table? Were there deals we'd win if we just had someone to pick up the phone? Maybe. Probably. But the question isn't whether SDRs would add pipeline. Of course they would. The question is whether the marginal pipeline justifies the cost, the management overhead, and the cultural shift of becoming a sales-driven organization before you've found product-market fit.