Your first real sales call will probably go badly, in the sense that you'll feel clumsy, say something twice, and hang up unsure whether it went well. That part is fine and it's not the part worth fixing. Plenty of founders who eventually get good at this were bad at it in exactly that way for their first dozen calls. What actually costs deals is a shorter, more specific list of habits, mostly ones technical founders arrive with by default, and every one of them is a behavior you can catch and change without turning into a smooth salesperson you don't recognize.
Answering the question they didn't ask
Someone asks what it costs. You explain the architecture. Someone asks if it works with their existing setup. You explain how the sync engine handles edge cases they haven't hit yet and might never hit. This habit comes from a good place, you understand the product deeply and want them to see that, but it reads to the person on the other end as avoidance, because you answered a harder, more interesting question than the one they actually asked. Answer the literal question first, in one sentence, and only then offer the detail, and only if they ask for it. The detail is usually for you, not for them.
Apologizing for the ask before you've made it
Watch your own messages for "no worries if not" and "totally understand if now's not a good time" attached to the end of an actual request. Each one quietly tells the reader that even you don't expect a yes, which makes a yes less likely simply because you signaled it wasn't really expected. You can be warm and low-pressure without undercutting your own ask in the same sentence. Say what you want plainly, and let their answer be their answer, instead of pre-writing an excuse for them to use.
Treating an objection as a bug report
Someone says the price feels high, or they're not sure this solves their exact problem, and the engineer in you hears a spec gap: build the feature, fix the pricing tier, patch the objection out of existence. Sometimes that's right. Often the objection isn't about the product at all, it's about trust, timing, or a decision they're not actually authorized to make alone, and no amount of feature work touches any of those. Before building anything in response to an objection, ask one more question: what would need to be true for this to be a yes. Half the time the honest answer has nothing to do with your roadmap.
Racing to fill silence
You ask a real question, and then, because the pause feels unbearable, you answer it yourself three seconds later, or soften it, or add five more sentences of context nobody requested. The pause was doing its job. People need a moment to actually think about what you asked, especially if it's a good question, and jumping in to rescue the silence usually rescues them from having to give you a real answer. Ask, then let it sit, even when it's uncomfortable. The discomfort is yours to carry, not theirs to be spared.
Dressing the price question up in the future tense
"So somewhere down the line, once you've had a chance to think it over, we could look at maybe figuring out pricing" is not a price question. It's five hedges wrapped around one, and it lets both of you avoid the actual moment where you say a number and they react to it. Say the number. Say it plainly, in the present tense, and watch their actual reaction instead of a hypothetical future one. You'll learn more from three seconds of their real face than from a paragraph of softened maybe-language, and you'll get to an actual answer faster.
Mistaking polish for the thing that builds trust
This is the one worth sitting with, because it runs against most sales advice you'll have absorbed by osmosis. A smooth, over-rehearsed pitch from an unknown company is not more trustworthy than an honest, slightly rough one. It's often less trustworthy, because sophisticated buyers have learned that polish is what a stranger reaches for when they've got nothing else, and a founder who says "I'm not totally sure, let me find out and get back to you today" reads as more credible than one who's never once admitted uncertainty on a call. You don't need to perform competence you don't have yet. You need to be honest about what's true, quickly, and follow through on what you said you'd check.
What "badly but working" actually looks like
It looks like a call where you stumbled over an explanation, asked a slightly clumsy question, and still walked away with a real answer, because you asked it plainly and let them answer. It does not look like a call where the product got over-explained, the price got softened into fog, and an objection got treated as a to-do item instead of a real signal. The difference between those two isn't polish. It's whether you did the handful of specific things above, not whether you sounded confident doing them.
The tell that you're improving isn't confidence, it's speed of recovery
Don't wait to feel confident before you decide you're getting better at this, because that feeling arrives much later than the actual skill does, if it arrives at all. Watch a different signal: how quickly you recover inside a call when something goes sideways. Early on, a hard question knocks you off track for the rest of the conversation. Later, the same hard question still lands, you still don't have a perfect answer, but you get back to asking real questions within thirty seconds instead of spiraling for the rest of the call. That recovery speed is the actual skill improving, quietly, underneath a feeling that may keep telling you you're bad at this for far longer than you actually are.
You don't need to become a salesperson to do this part
None of the above requires the personality type usually associated with selling. It requires showing up for the reply, asking directly, and sitting still through the pause. The mechanical half, finding the right person to call, writing the first message, catching the ones who call back instead of writing, can run without you: Hand Rocketship a plain-English description of who you're trying to reach, and it will find those people, write to them from your own inbox, and pick up your phone on the days someone calls before you've even finished the email. That leaves exactly the part only you can do, and it's a shorter list than it feels like from the inside: ask the direct question, say the real number, and let the silence sit.