Cold calls still work when the script is sharp, short, and respects the fact that the prospect did not ask to be interrupted. The scripts that fail are the ones built around company talking points instead of buyer context. The scripts that succeed open with a reason the call might matter, surface one specific pain in two questions, deliver one credible proof point, handle the top objection in a sentence, and ask for a specific next step. This guide walks you through the full build: researching the account, writing the 30-second opener, scripting two anchor discovery questions, drafting a single proof point, handling the top five objections, and closing with a calendar hold within the same call.
Time: 60 minutes
A cold call script that does not reference anything specific about the account sounds like a telemarketer reading. Research takes two minutes per account and transforms the first 30 seconds of the call. Pull the account from the CRM, check recent news (funding round, exec hire, product launch, layoff), look at the prospect's LinkedIn for role tenure and recent posts, and check your enrichment data for firmographic signals (headcount growth, revenue tier, tech stack). From the research, pick one specific hook that gives the prospect a reason the call might be worth 30 seconds. The hook could be a recent event (I saw the funding announcement), a role transition (congrats on the new role two months in), or a pattern (companies in your segment at your size often hit a specific challenge). The research is the fuel for the opener; without it, the opener defaults to generic company pitching that gets hung up on in six seconds.
The opener has one job: earn the next 30 seconds. The structure that works: name the prospect, name yourself and your company in one short line, reference the hook, ask permission to continue. Example: "Hi Priya, this is Marcus from Strkr. I saw your Series B announcement last month and figured you are probably rethinking the sales stack. Mind if I ask one question, then you can tell me if this is worth a longer conversation?" The opener is 25 to 30 seconds spoken. It names the prospect by first name, states who is calling without burying it, references the specific hook from the research, and asks permission with a specific framing that promises a short interaction. Avoid vague permission asks like "do you have a minute to talk." Prospects say no automatically. "Mind if I ask one question" lands differently because the ask is specific and finite. Rehearse the opener aloud until it sounds conversational, not read.
Cold calls do not have time for ten discovery questions. Pick two anchor questions that qualify the account and surface a real pain in under two minutes. Question one is about the current state and the trigger: "How are you handling the pipeline forecasting today, and what prompted you to even consider changing?" The question invites a direct answer, surfaces whether a trigger exists, and lets the prospect say "nothing prompted me, we are fine" (which is itself qualifying information). Question two is about the cost of the current state: "What does the current approach cost you in time or missed deals?" The question moves the conversation from status quo to pain. If neither question surfaces a real pain or trigger, the cold call is probably over, and the right move is to disqualify respectfully and ask for a referral to the person on the team who handles the specific pain. Two questions, two minutes, enough signal to decide whether the call deserves more time.
After discovery, deliver one proof point that connects your product to the specific pain the prospect just named. The proof point is a one-sentence customer story with a specific number: "We work with a mid-market SaaS team your size that cut forecast variance from 25 percent to 8 percent in a quarter." The sentence has to be specific (named segment, size, metric, timeframe) and credible (not aspirational marketing numbers that reps will not vouch for). Follow the proof point with a transition question: "Does that resonate with what you are seeing?" The question invites the prospect to react, which keeps the call interactive rather than a one-sided pitch. Avoid listing features, running down the product tour, or reciting the marketing deck. A single credible proof point tied to a specific pain is more persuasive than a 90-second overview, because it lets the prospect place themselves in the story.
Cold calls produce the same five objections most of the time: not interested, send me info, we already have a solution, call me back, we have no budget. Script a one-sentence response to each, built around acknowledge-reframe-ask. Example for "we already have a solution": "Makes sense, most teams your size do. The question I usually ask is whether the current solution is cutting forecast variance to under 10 percent, because that is where most teams stall. Where does yours land?" The response acknowledges the status quo, reframes to a specific metric, and asks a question that advances the call. Keep every response under 15 seconds spoken. Longer responses sound defensive. Store the responses in the live-call playbook surface so reps can see them during the call, not in a training deck. The responses evolve: as prospects respond to one set of lines, prepare the next set, and tune every month from recorded calls.
The close is where a cold call either produces a calendar invite or evaporates. Avoid vague closes like "want to learn more" or "I'll send some info." Both are easy to dismiss and dead within 48 hours. The close that works names a specific calendar hold on the current call: "What I'd suggest is we grab 20 minutes next Tuesday or Thursday to walk through what we did with that team. Do either work?" The ask offers two specific days, respects the prospect's time with a 20-minute window, and ties the meeting to the proof point just referenced. If the prospect accepts, send the calendar invite within five minutes of hanging up. If the prospect pushes back on scheduling, offer an asynchronous alternative: a two-paragraph email summary and a short video demo. The async option saves the lead for a later call, but every cold call should push for the synchronous booking first because async conversion rates are a fraction of booked meetings.
Most cold calls hit voicemail. The voicemail variant is a 20-second message with a specific hook and a specific callback ask: "Hi Priya, Marcus from Strkr. Short reason I called: we worked with a mid-market SaaS team that cut forecast variance from 25 percent to 8 percent. If this is a problem on your plate, I'll follow up by email, or you can grab my calendar at [short link]. Thanks." The voicemail references the same hook as the opener and gives one specific call to action. The callback variant covers the case where the prospect calls back and the rep needs to pick up without the research in front of them. Script a shorter opener that acknowledges the callback and asks a direct discovery question. Store both variants in the live-call playbook alongside the main script so reps have the right language in both situations.
A cold call that lands on voicemail or ends without a booking should trigger a follow-up email sequence. Write three emails: email one goes out the same day with a two-paragraph summary of the pitch and the calendar link; email two goes out 48 hours later with a different angle (a specific customer story, a short piece of content); email three goes out five business days later as a breakup email offering to close the loop if the timing is wrong. The three-email sequence respects the prospect's inbox while keeping the lead warm. Tie the sequence to the CRM so it fires automatically after a cold call without a booking. Monitor open and reply rates on each email; patterns in which email produces the booked meetings tell you which hook, story, or framing is landing. Over time, the data from the follow-up sequence often improves the cold call script itself.
Before rolling the script to the full team, pilot it with two SDRs for ten business days. Pick one top SDR and one average SDR so you see how the script performs across skill levels. Have both record every call, review ten calls together each week, and track three numbers: contact rate (calls that reached a human), conversation rate (contacts that produced a discovery exchange), and booking rate (conversations that produced a scheduled meeting). The pilot surfaces where the script lands and where it stalls. Expect to revise. Lines that sound unnatural on the page sometimes read fine until the prospect responds in a way the script did not anticipate. Rewrite in week one based on week-one data, run week two on the revised version, and decide at end of week two whether to roll out, revise further, or revert. Short pilots produce enough signal to decide; long pilots invite endless rewrites that never ship.
A script dropped into a shared drive is a script nobody uses. Pair rollout with structured coaching: a one-hour team session walking through every section of the script with examples from recorded calls, a shadow program where every SDR listens to a top SDR run the script three times live, and a weekly call review cadence for the first four weeks focused on where SDRs are following the script and where they are drifting. Expect drift. SDRs will shortcut the opener, skip the second discovery question, and improvise the close. The weekly review surfaces where the drift is helping (good improvisation to roll back into the script) versus hurting (bad improvisation to correct in coaching). Store the script in the live-call playbook surface so SDRs see it during the call, not in a wiki page they visit once. Reps follow scripts they can see live and that produce measurable wins in the review.
Cold call scripts age. The market shifts, buyers develop immunity to specific openers, proof points land differently as competitors move, and objection patterns evolve. Every month, pull a sample of 20 recorded cold calls, listen for patterns the current script does not handle well (a new objection appearing repeatedly, a hook that stopped working, a close that lands flat), and revise. Keep a version history so you can roll back if a revision underperforms. Watch the three pilot metrics (contact rate, conversation rate, booking rate) month over month; a healthy script holds steady or improves on all three. Pair the monthly revision with a quarterly deep review where you reread the full script, compare it to current recorded calls, and prune any line that no longer earns its place. Pruning is as important as adding. Scripts that only grow turn into unmaintainable documents reps work around.
Short enough to fit the full flow on one page. Opener (30 seconds), discovery (two minutes), proof point (30 seconds), objection responses (one sentence each), close (one minute). Total script length on the page should be well under 500 words, because longer scripts become unreadable mid-call when the SDR has to improvise.
The structure should be the same, but the exact wording can flex. Give every SDR the same five-part flow (opener, discovery, proof point, objection handling, close), the same anchor questions, and the same proof points. Let SDRs adjust the surface language to match their voice. Top SDRs who sound natural outperform SDRs who sound like they are reading, every time.
Five is the right number. Cover the top five objections your team hears repeatedly (not interested, send me info, we already have a solution, call me back, no budget) with one-sentence responses. Any more and the script becomes unreadable. Any fewer and SDRs are improvising on the common objections they should have scripted answers for.
Varies by segment and territory. For mid-market SaaS cold calls, a healthy script produces 5 to 10 percent booking rate on connected calls (calls that reached a human and had a conversation). SMB can run higher at 10 to 15 percent; enterprise typically runs lower at 3 to 7 percent. Benchmark against your own historical data, not vendor averages.
Yes. A voicemail variant should be 20 seconds, reference the same hook as the opener, and give one specific call to action (a calendar link or an email follow-up promise). Script the voicemail variant alongside the main script so SDRs have the right language ready in the moment, not improvising on the fly.
Use Strkr AI to analyze recorded calls, surface patterns, and suggest revisions, but write the script yourself. AI-generated scripts tend toward generic language that sounds like marketing collateral; scripts rooted in your top reps' actual recorded calls produce more bookings. Strkr AI is best deployed on the analysis and iteration side, not the drafting side.
Strkr pins the live script to the dialer, surfaces objection responses in context, and feeds recorded-call data back into the next revision. One source of truth, zero playbooks collecting dust.
No sales call, no migration consultant, no four-month implementation. Enter your card, get 14 days of the full Pro tier, cancel any time before day 14 with zero charge. Spin up a workspace, import your CSV, and have something useful before lunch.
