Context

Christina Gilbert, a YC visiting partner who personally generated and closed millions of dollars through founder-led sales at OneSchema, breaks down eight concrete ways early-stage founders can fix outbound sales that isn't getting replies. This matters to PMs at early-stage companies because founder-led outbound is really a discovery and positioning exercise in disguise: who has the problem, what language makes them respond, and how you turn their replies into sharper messaging. The tactics translate directly to how a PM validates a target segment and a value proposition before scaling either.

The Big Idea

Targeting who you reach out to matters more than how well you write the message, and the fastest way to learn both is to do the first 100 outreaches by hand, not through automation.

Founders who automate outbound before learning anything about their audience get zero replies and learn nothing from it, because the failure could be messaging, targeting, deliverability, or the subject line, with no way to tell which. Manual, individualized outreach is what actually surfaces the winning positioning.

Key Insights

Automating before learning teaches you nothing

The most common mistake Gilbert sees in office hours is a founder who has already automated a campaign, blasted out a thousand emails, and gotten zero replies. The problem is that a zero-reply automated blast is undiagnosable: the cause could be messaging, company targeting, prospect selection, deliverability, or the subject line, and there's no signal to isolate which. Her fix is to do at least the first 100 outreaches by hand, researching one person's specific problem and what would actually catch their attention, because that manual homework is what produces the highest-converting outbound ideas.

Wrong job titles doom even a great message

Gilbert worked with a winter-batch founder selling an infrastructure product meant for senior engineering leaders, but whose outreach list was full of DevRel managers, GTM managers, and even a customer success manager, none of whom had the pain point the product solved. No amount of message polish would have fixed this, since those job titles were never going to buy regardless of how the email was written. She finds the right targets by tracing existing customers' sales processes (who pushed the deal to close, who signed, who paid) and reaching out to people with similar profiles, or, pre-customer, by asking who in an org has the most to gain from the product, using the example that a chief security officer (who could get fired over a breach) is a much better target for device-monitoring software than an engineering manager (who has no upside from being monitored).

List-building tools return the wrong companies too

A search for "tech companies" in a tool like LinkedIn Sales Navigator often surfaces a batch of IT implementation consultancies that only loosely match the query, not real targets. Gilbert's fix is to individually verify that each company in a generated list is actually a fit before sending, since an unverified list means the message falls on deaf ears no matter how well it's written.

Intent signals tell you when a company is ready

Job postings and company growth stage are two of the clearest intent signals Gilbert points to; a compensation-planning tool, for instance, isn't needed by companies under 100 employees (too early for formal levels) or by companies over 500 (already have bands set), so the signal to reach out sits in that crossing-100-employees window. Every product has its own version of this signal, and Gilbert frames finding it as a founder's job, not something a generic list-building tool will hand you.

The best cold email explains almost nothing about the product

Gilbert breaks down a real email from an S24 YC founder (Apten) that she considers a model: it opens by establishing credibility through a relevant past role ("I used to be a software engineer at Tesla. We got crazy good at using SMS to convert the people who visited our website into car buyers"), personalizes to something specific the founder noticed on the prospect's site, and ends with one crisp, specific ask ("would you be open to an experiment to see if we could improve your conversion rate"). It barely explains how the product works at all; the entire email is built around demonstrating that the sender understands the buyer's problem, not the product's features.

Executives buy outcomes, not features

Buyers don't care that a product has fast uploads or is built in Rust; they care about hitting their quarterly target, avoiding getting fired, or getting promoted. Gilbert's rule is to compress the product explanation to one sentence and spend the rest of the email establishing why the sender is credible to help, saving feature detail for the first call rather than the cold email.

A specific ask beats a vague one

"Got 20 minutes to hop on a call this week?" converts better than "does this sound interesting to you?" because it removes the buyer's need to interpret what happens next; pairing it with a calendar link or a few concrete time slots removes even more friction from replying.

Mental Models & Frameworks

The six-layer outbound debugging order

When a founder is getting a low but nonzero reply rate (Gilbert's baseline for "you have something to iterate on" is 2 to 3 replies per 100 emails), she debugs in this fixed order rather than jumping straight to rewriting copy:

• Right person: are you emailing someone who actually has the pain point?
• Right companies: are the companies on your list genuinely a fit, not just keyword matches?
• Subject lines: is the subject (and the first 20 to 30 characters of the body, which shows in the Gmail preview) getting the email opened at all?
• Messaging: only after the first three are solid does she look at the actual email copy.
• Materials: does the website and LinkedIn profile hold up once someone clicks through?
• Deliverability: is the email even landing in the inbox?

If all six check out and replies are still at zero, Gilbert treats that as a signal worth investigating for a deeper product-market-fit problem, not just an outbound problem.

LinkedIn profile as a landing page

Gilbert treats a LinkedIn profile the way a PM would treat a landing page: the banner should function as a concise, jargon-free ad for the company, the about section should be rewritten as a short description rather than a life story, and anything irrelevant to a buyer (hobbies, unrelated experience) should be cut so the relevant, credibility-building details are what a visitor actually sees. Building relevant mutual connections matters too, since shared contacts are, in her experience, the single biggest predictor of whether someone accepts a connection request.

Practical Application

Do your first 100 outreaches by hand

Before loading any outreach into an automation tool, personally research and message at least 100 individual prospects, reading their LinkedIn and thinking through their specific problem for each one. Treat this as market research, not a sales shortcut: the goal is to learn which targeting and phrasing actually gets replies before scaling either.

Trace your best customers' buying path to define your target profile

Look at the customers you already have (or, pre-revenue, ask who in a target org has the most to lose or gain from the problem you solve), identify the job titles of whoever pushed the deal to close, signed the contract, and paid, and build your outreach list around people who match those profiles rather than a generic keyword search.

Compress your product pitch to one sentence

Rewrite your cold outreach so the entire product explanation fits in a single sentence focused on the buyer's outcome (what keeps them up at night, what would get them promoted or fired), and move all feature detail to the first call instead of the initial message.

Run a fixed 4-follow-up sequence, ending with a breakup email

Space 2 to 4 follow-ups a few days apart rather than sending one message and giving up, and make the last one a low-pressure "breakup" message (something like "it seems like this isn't a priority for you right now, just let me know and I'll spare your inbox"). Gilbert finds these get disproportionately high reply rates because they make saying no easy, and a real no is still useful data for refining future targeting.

Ask every new customer what caught their attention

On your very next sales call, ask "what got your attention about my message?" and separately ask existing customers the single thing that got them to buy, then reuse their actual phrasing in future outreach. Customers consistently write better copy about their own problem than a founder can guess at from the outside.

Bottom Line

Reply rates are a targeting and learning problem before they're a copywriting problem: manually researching the first 100 prospects, verifying you're reaching the right job titles and companies, and feeding customer language back into your messaging will move outbound further than polishing the email itself.