Column: A sales reply is information, not applause

The most dangerous reply to a founder’s sales email may be a friendly one. Somebody says the product looks interesting, offers encouragement and agrees to stay in touch. The founder records momentum. The buyer returns to work. Nothing dishonest has happened, yet both sides can leave the conversation with entirely different ideas about what it meant. I think early sales improves when founders treat replies as evidence to interpret, rather than applause to collect.

In Y Combinator’s 12 September episode featuring Christina Gilbert, the publisher’s description recommends beginning outreach manually and learning who has the problem before automating. That advice creates a useful opening for a narrower question: what, exactly, should a founder learn from a reply? The following is my interpretation of that challenge, not a reconstruction of the full episode. The recording’s title was updated after publication; the underlying video is the same source.

Courtesy is not a purchase process

Imagine a founder offering a tool that helps small agencies collect client feedback. An agency owner replies that the idea sounds useful. That tells the founder the message was understandable enough to receive a response. It does not establish that collecting feedback is a priority, that the owner has a budget or that the proposed solution fits the agency’s workflow. These are separate questions. A pleasant reply cannot answer all of them at once.

The next conversation should become more specific without becoming an interrogation. Ask how the agency handled feedback on its most recent project. Where did information get lost? Who chased the client? What happened when the response arrived late? A real example gives the founder something to investigate. A hypothetical promise to use a future product gives much less. This distinction also helps keep the conversation respectful: the buyer can describe work rather than defend enthusiasm.

The U.S. Small Business Administration’s market research guidance separates questions such as demand, market size and alternatives. A founder’s inbox needs similar distinctions. One response may reveal a language problem; another may reveal that the wrong person received the message. Neither automatically says the product is bad.

Keep a record of what changed

I would rather see a small learning log than a triumphant chart of messages sent. After a conversation, write down one assumption that became stronger, weaker or still unresolved. Perhaps the pain belongs to the project manager rather than the owner. Perhaps it occurs only with larger clients. Perhaps the agency already solves it through a service the founder had not considered. These findings should change the next question or the next target.

This is a useful companion to the argument for choosing a narrower niche. A niche is not simply a demographic label. It can be a recognisable working situation in which a problem becomes urgent. Two businesses in different industries may share that situation, while two businesses in the same industry may not. Better targeting comes from understanding that difference, not from adding more decorative personalisation to the opening sentence.

There is a temptation to polish the message after every rejection. Sometimes that helps. Sometimes it disguises a weak offer. If the customer understands the proposal and still has no reason to act, changing the subject line may produce more conversations without improving the business. Founders need room to discover that the problem is less important than expected. A learning process that can only recommend more outreach is not really open to evidence.

Automation should preserve the lesson

None of this makes automation the enemy. Once a founder understands a recurring task, software can remove repetition and make follow-up more reliable. The question is whether the system preserves the distinction between different responses. A request for information, a referral to a colleague and a refusal should not become three identical entries in a generic engagement score. They imply different next actions, including sometimes no further action at all.

The material collected in sales and marketing channels can help founders compare approaches, but copying somebody else’s cadence cannot replace judgement about their own buyers. A busy procurement manager and a sole trader do not necessarily respond to the same rhythm. The founder should be able to explain why the next contact is useful to the recipient. “The sequence says so” is a description of a system, not a reason.

A small team should also agree what counts as progress. A scheduled conversation is progress towards learning. A confirmed problem is progress towards qualification. An agreed evaluation is progress towards a buying decision. These stages should not be collapsed into a single success label. Keeping them distinct makes disappointing results easier to discuss because the team can locate the actual gap instead of blaming the person who wrote the email.

Let the customer correct the story

One of the best uses of a sales conversation is discovering that your explanation is wrong. The founder calls the benefit efficiency; the buyer calls it avoiding an embarrassing handover. The founder describes a dashboard; the buyer wants somebody to tell them which issue needs attention. These differences can change positioning, onboarding and the product itself. They are difficult to hear when the seller is busy defending a prepared narrative.

That is why founder and startup discussions are most useful when treated as prompts for investigation. Another company’s success story can suggest a question, but it cannot supply your customer’s answer. Early sales is tiring partly because the uncertainty is real. Pretending it has disappeared through a well-designed campaign dashboard does not make the company more mature.

A reply is an opening to learn what happens inside another business. Sometimes it leads to a sale; sometimes it leads to a better question or a decision to stop pursuing that segment. All three can be valuable. I would count fewer compliments and pay closer attention to the moments when a buyer changes the founder’s understanding. That is the information worth carrying into the next hundred messages.

▶

Source: Y Combinator, Christina Gilbert, 12 September 2026. The publisher’s description was reviewed; no claim is made to have watched the complete episode.