Cold call script: build your own in 6 steps

Most cold call scripts you find online are someone else's words. You download a finished script from a vendor, read it in, and sound like the thousand other reps who downloaded the same file. The problem isn't that the script is badly written. The problem is that it was written for a different product, a different buyer and a different voice than yours.

This is a guide to the craft, not a template to copy. We walk through the six parts a cold call script is made of and what each one is meant to accomplish. Once you understand the function you can write your own lines, and more importantly, swap them live without losing the thread. Salesprep has a cold call module where you can run your finished script ten times against a voice AI that plays the customer before you dial for real, with automatic feedback on opener, structure and objections so you can see what actually holds up.

Why build your own script instead of copying one?

A copied script ties you to the words. When the buyer says something unexpected, which they always do, you have no idea which part you're in, so you just keep reading and sound like a recording. Your own script ties you to the function instead. You know you're in the opener, that the next step is the reason for the call, and that the end goal is the meeting. That lets you respond to what the buyer actually said and still find your way back to the structure.

This isn't an argument against scripts, it's an argument against rigid reading. In Cognism's 2024 data, 52 percent of reps use a script, but they adapt it heavily during the call. It's the difference between a pianist who knows the piece and one reading sheet music for the first time. Both have something on the stand. Only one can listen to the room at the same time.

Step 1: What research do you need before you pick up the phone?

Before you write a single line you need to know who you're calling and why now. Two things are enough: a persona and a trigger. The persona is the role, a sales manager has different problems than a CEO. The trigger is something concrete and recent that happened at the account: a new hire, a new product, a funding round, a change in leadership. The trigger is what keeps the call from sounding like the next name on a list, and it gives you a natural reason to call today specifically.

You don't need an hour of research per contact. You need the role, a relevant trigger, and a guess at what problem that trigger creates. That guess becomes your value hypothesis in step 4. Everything else is nice to have, but it isn't what decides whether the call turns into a meeting.

Step 2: How do you open so they don't hang up?

The opener has three parts: your full name, your company, and a pattern interrupt. Gong analyzed 90,380 openers and found that leading with full name and company, then breaking the pattern, performs measurably better. The single strongest one they found was a simple courtesy: "How have you been?" produced 10.01 percent success against a 1.5 percent baseline, 6.6 times better. It works because it sounds human in the middle of a call the buyer expected to be robotic.

One opener to avoid: "Did I catch you at a bad time?" It produced 0.9 percent success in Gong's data and made the call 40 percent less likely to lead to a meeting. You hand the buyer a ready-made exit before you've even said why you're calling. Break the pattern, but break it in the right direction, toward a question that opens the call rather than closing it.

Step 3: What do you say right after the opener?

Say why you're calling, plainly. The line "The reason for my call is..." produced a 2.1x lift in Gong's data. It sounds almost too simple, but it does two things at once: it gives the buyer a frame for what's coming, and it forces you to have an actual reason. If you can't finish that sentence in a way that's relevant to this specific person, you aren't ready to call yet.

Tie the reason to the trigger from step 1. "The reason for my call is that I saw you've hired four reps in the last month." Now the opener and the reason are one unit: a pattern interrupt followed by a concrete, personal reason. This is where most copied scripts fall apart, because their reason is generic and fits no one in particular.

Step 4: How do you write a value hypothesis?

A value hypothesis is an educated guess about a problem the buyer likely has, not a claim about your product. It comes from the persona and the trigger: "We usually see that the period right after a hiring wave is when coaching time gets the most stretched." Notice you aren't selling anything. You're putting forward a hypothesis the buyer either recognizes or corrects you on, and both answers are valuable. If they recognize it you have a conversation. If they correct you, you've learned more in five seconds than a monologue would have given you.

The value hypothesis should be about the buyer's world, not your feature. The difference is small in words but decisive in tone. "You probably struggle to coach new reps" is a hypothesis. "Our platform coaches new reps" is a pitch. The first invites dialogue, the second signals you've already decided what they need.

Step 5: How do you move from script to dialogue?

After the value hypothesis you need a transition that hands over the floor. The simplest is a short question that checks the hypothesis: "Does that hold true for you?" or "How does it look on your end?" This is where the read part of the script ends and the call really begins. This transition is not a full discovery, it's a door. You ask one, maybe two questions, just enough to decide whether there's a reason to meet, not to solve the whole deal on the phone.

The most common mistake here is trying to run the entire discovery on the cold call. You find a problem, get excited, and start digging into details, budget and decision process. Now you've turned a two-minute call into a meeting the buyer never agreed to. Note the interest, ask your transition question, and move to the one thing the call is actually about.

Step 6: How do you ask for the meeting?

The goal of a cold call is the meeting, not the discovery. That's the single most important thing to understand about the whole structure. Everything you built, the opener, the reason, the hypothesis, the transition, exists to make one yes possible: yes to a booked meeting. So ask for it directly, with a concrete proposal: "Do you have twenty minutes Thursday morning and I'll walk you through how others in your situation solved it?" A specific proposal is easier to answer than a vague "maybe we could meet sometime."

This frame also explains why call length is a weak metric on its own. Gong found that successful cold calls run about twice as long as unsuccessful ones, with an ideal somewhere around five minutes and fifty seconds. But length is a consequence, not a target. The call runs longer because the buyer is interested enough to stay, not because you talked more. Build for the interest and the time follows.

Put the parts together and rehearse the script

Now you have six parts: research, opener, reason, value hypothesis, transition and the ask. Write them out as a single page, a sentence or two per part, in your own words and your product. That's your script. But a script that has only lived on paper always sounds like paper in the ear the first few times you say it out loud, so the last step is rehearsal.

Run the script ten times in Salesprep's cold call module against a voice AI that plays the buyer, before your first real dial. You get automatic feedback on opener, structure and how you handle objections, and you can call the same stressed CEO over and over until the transitions stick. When the parts feel like your own words and not a document, you're ready to pick up the phone for real.