Outreach that resonates from the first line

Paste a draft and hear how the buyer reads it. Then send the rewrite in your customers' words, with a proof point from a deal you won.

Start free trial

Works inand any MCP client

Question: Would Marcus Lee at Ledgerline reply to this? If not, rewrite it: "Hi Marcus, our AI-powered platform gives you end-to-end observability across your data stack. Got 15 minutes next week?"

Used Calven MCP

Persona verdict: deleted. The Head of Data stops at "AI-powered": every vendor in his inbox says it, and nothing in the email names a problem he has.

Rewrite, in your customers' words: "Hi Marcus, most data teams we talk to hear about a broken pipeline from finance before their own tools catch it. Kestrel Payments fixed that in six weeks. Worth comparing notes on how Ledgerline catches it?"

Verdict on the rewrite: names his problem and a proof point. Worth sending.

Sources: Head of Data persona review · 11 customer calls · Kestrel win story

Example workspace. Names and numbers are illustrative.

What can BDR and SDR teams use Calven for?

Emails, call scripts, sequences and account briefs that start from the approved story and the words your buyers use on calls. Calven is the approved context your AI workflows need to scale across the team and still be trusted.

Use cases

Example prompts for outreach drafts

Ask any of these in Claude, ChatGPT or any AI tool connected to Calven.

Pick the three cold email openers [persona] in [segment] is most likely to answer.
Would [contact] at [account] reply to this? If not, rewrite it: [paste your draft]
Can I say this in an email? Check it against our sources: [paste the proof line]
Tell me why email A beat email B with [persona]: [paste both, with reply rates]
Drop my email into a simulated [persona] inbox next to rival emails and see if it survives a three-second triage.
Delete my email one sentence at a time and see which sentences [persona] would miss, so I cut the dead weight.
Score my email against the words [persona]s actually use on calls, and swap our jargon for their phrases.
Would [persona] reply to this: "[paste one line]"?
Give me the three phrases customers use for [pain].
Give me a subject line under 40 characters about [pain], in the customer's words.

Open the full prompts on GitHub

Example prompts for account research

Ask any of these in Claude, ChatGPT or any AI tool connected to Calven.

Brief me on [account] before my first touch: fit, who to contact, what to say.
I can send three messages at [account] this week. Who gets them, and in what order?
Find the won deal closest to [account] in [industry] and tell me what decided it.
Have we worked [account] before? Who did we talk to, and how did it end?
Estimate [account]'s real odds of becoming a won deal from our base rates, attribute by attribute, before I spend an hour on it.
Map who at [account] has power and who's likely on our side, then plan which door to knock on first.
Argue [account] out of my list first: hunt for every reason it won't buy before I write a word.
s ICP fit and why?",
,
Which attribute of [account] lowers its odds most, according to the ICP dashboard?

Open the full prompts on GitHub

Example prompts for brush-off answers

Ask any of these in Claude, ChatGPT or any AI tool connected to Calven.

Which brush-offs from [persona] in [segment] turned into won deals, and what changed?
A [persona] said "we already use [competitor]". Give me two lines from buyers we won.
Write two-line answers to the brush-offs I hear most from [persona]: [list them]
Play [persona] on a cold call about [product] and open with [brush-off].
Run a Delphi panel of our personas and evidence to sort which brush-offs mean no and which mean later.
Replay deals we lost to [competitor] as the incumbent, and test whether a different first answer would have changed them.
Plot every brush-off on how often [persona] says it and how often we still win, and practise the top-right corner.
How do we answer "we already use [competitor]"?
Which brush-off do [persona]s give most often?
Give me a quote from a customer who switched from [competitor].

Open the full prompts on GitHub

Example prompts for sequence reviews

Ask any of these in Claude, ChatGPT or any AI tool connected to Calven.

Review this sequence for [persona] in [segment] before it goes live: [paste it]
Check this sequence against prospects who already run [competitor]: [paste it]
What is this [persona] sequence missing that buyers we won needed to hear?
Rewrite touch [3] for [persona] in the words customers use on calls.
Model my sequence as a chain of touches, find where prospects leak, and size the pipeline a fix to that touch would recover.
Hand my sequence to [competitor]'s SDR lead and have them plan how to beat it in the same inbox.
Write a scored rubric and a graded example set, so every future sequence gets checked the same way before launch.
Is this sequence on-message for [persona]? [paste]
Which objection does this sequence never answer?
Rank the five touches by how likely [persona] is to reply.

Open the full prompts on GitHub

Example prompts for buying signal plays

Ask any of these in Claude, ChatGPT or any AI tool connected to Calven.

Which triggers came before the deals we won in the [last 12 months]?
List the [20] in-profile accounts where [trigger] fired, with the [persona] to contact.
Write a three-touch play for [persona] at [segment] accounts right after [trigger].
Which of these accounts have no [persona] on record? [paste the list]
Measure the real lift each trigger gives our win rate, with confidence intervals, before I build plays around it.
War-game a funding trigger against [competitor]'s SDRs: who moves first, with what angle, and who wins the meeting.
Model trigger plays as a queue: how many my week can hold and which ones to let go cold.
Which trigger shows up most on accounts we won?
Which accounts had a funding trigger recorded this quarter and are Tier 1?
Which trigger appears on our lost deals as often as on our won ones?

Open the full prompts on GitHub

Example prompts for target list building

Ask any of these in Claude, ChatGPT or any AI tool connected to Calven.

Write targeting rules for [segment] next quarter from the deals we won.
Build a ranked list of [50] accounts in [segment] for [next quarter].
Give me one reason to call [persona] at each of these accounts: [paste the list]
Which buying roles are we missing at each of these accounts? [paste the list]
Fit a win model on our closed deals and score every unworked account in my territory with it, with a holdout check.
Split my quarter's touches across tiers and segments like a portfolio, to get the most expected pipeline for my hours.
Run a Monte Carlo on my target list and tell me the realistic pipeline range it produces this quarter.
Which Tier 1 accounts in [industry] have never had a deal?
What attributes predict a win for us?
Which accounts have a Tier 1 fit but no contact on record?

Open the full prompts on GitHub

Example prompts for win-back outreach

Ask any of these in Claude, ChatGPT or any AI tool connected to Calven.

List the deals we lost on [loss reason] that [what changed] now answers.
Write the win-back message for [account], lost in [month] over [reason].
What changed at [competitor] since we lost [account] to them?
Confirm [capability] is in the product before I tell lost accounts about it.
Score every closed-lost deal on reopenability with a weighted rubric, then check which ranking survives when the weights move.
War-game [competitor]'s response when I try to win back [account]: their retention move, my counter, and who blinks.
Run a survival analysis on our closed-lost accounts to see when they typically come back, so I time win-backs right.
Which deals did we lose on "Missing feature" in the last year?
Is the champion from [account] still a contact?
Who did we lose [account] to, and what did the buyer say?

Open the full prompts on GitHub

Example prompts for call scripts

Ask any of these in Claude, ChatGPT or any AI tool connected to Calven.

Build a cold-call talk track for [persona] in [segment], with the three objections.
How did our reps open the calls we won with [persona]? Quote them.
Fact-check every claim in this talk track: [paste the talk track]
Play [persona] and let me rehearse a cold call about [product].
Build me a clickable call flow for [persona]: one screen, every branch, the approved answer one click away.
Pit twelve openers against each other in a tournament judged by [persona], and keep the two with the highest rating.
Feed in my dialer export and find which openers, personas and times actually drive booked meetings, controlling for the rest.
Give me a fifteen-second opener for [persona] on [pain], in the customer's words.
Which objection does [persona] raise most, and what is the approved response?
What should I never say to [persona] in the first minute?

Open the full prompts on GitHub

Example prompts for AE handoff notes

Ask any of these in Claude, ChatGPT or any AI tool connected to Calven.

Turn my notes on [contact] at [account] into a [MEDDIC] handoff note: [paste notes]
Is [account] worth an AE meeting, based on [contact], [title]?
Brief the AE on [competitor] at [account]: where we win and what to avoid.
Backtest last year's handoffs: which facts at handoff actually predicted a won deal, so my notes lead with those.
Put pass, nurture and drop side by side as an expected-value tree for [account], using our real win rates and deal sizes.
Is [account] in profile?
What buying role is a [title] usually in our won deals?
Which disqualifiers should I check before handing this off?
Which contact role on a handoff most often ends in a won deal for us?
What do lost deals in [segment] have in common at the first meeting?

Open the full prompts on GitHub

Example prompts for inbound replies

Ask any of these in Claude, ChatGPT or any AI tool connected to Calven.

Draft a reply to [contact] at [account], who asked: [paste the inbound message]
A [persona] asked how we compare to [competitor]. Draft the honest answer.
Is a [size] [industry] company in [region] a fit for us?
Work out what each extra hour of inbound response time costs us in pipeline, and where adding coverage breaks even.
Grow a qualification decision tree from our closed inbound deals, so every SDR asks the two questions that matter.
Does our product do [capability]?
How do we compare to [competitor] on [capability]?
What is the right next step for a [persona] at the evaluation stage?
What's our win rate on inbound deals compared to outbound, and on what n?
Which persona asks for pricing first, and what does our messaging say back?

Open the full prompts on GitHub

Example prompts for LinkedIn messages

Ask any of these in Claude, ChatGPT or any AI tool connected to Calven.

Write a connection note and two messages for [persona] at [account].
Read these as [persona] and tell me which one gets a reply: [paste them]
My first message to [persona] was about [pain]. Give me a new angle.
Treat my LinkedIn angles as a multi-armed bandit: update on every reply and shift messages to the angle that's winning.
Cross my LinkedIn connections with contacts on our won deals to find warm introductions into [segment] targets.
Give me a 300-character connection note for a [persona], no ask.
Is this message a pitch or a problem? [paste]
Which customer story would a [persona] in [industry] recognise?
Which watering holes on [persona]'s canvas are LinkedIn groups or creators?
Which customer quote on [pain] fits in 300 characters?

Open the full prompts on GitHub

Example prompts for event follow-ups

Ask any of these in Claude, ChatGPT or any AI tool connected to Calven.

Write the [event] follow-ups for [persona]: attended, visited the booth, no-show.
Rank these event leads by who to call first: [paste the attendee list]
Find the phrases customers use about [theme] for my event follow-ups.
Fit a lead score on my last event's leads and outcomes, then use it to rank this event's list before I call anyone.
Run a conjoint-style test on my follow-up: which offer, proof and ask [persona] values most, played by our personas.
Which of these attendees are at Tier 1 accounts? [paste]
Give me a no-show follow-up for [persona], under 60 words.
Which phrases did customers use about [theme]?
Which persona converts to a deal most often after a first meeting, per the persona dashboard?
Which of these companies were lost deals, and to whom? [paste]

Open the full prompts on GitHub

Example prompts for voicemail scripts

Ask any of these in Claude, ChatGPT or any AI tool connected to Calven.

Write two voicemails under 60 words for [persona] in [segment].
Give me one proof line for [pillar] that I can say aloud to [persona].
Would [persona] call back after this voicemail? [paste the voicemail]
Play my voicemail to a simulated [persona] once and see what they remember an hour later, then rewrite for recall.
Design a 2x2x2 voicemail test so I learn which of three choices drives callbacks from one month of dials.
What is the one pain [persona] would stop for in a voicemail?
Give me a 50-word voicemail for [persona] on [pain].
What should I never say in a voicemail to [persona]?
Which customer quote is short enough to say in a voicemail for [pain]?
What's one number from our proof points a [persona] would remember?

Open the full prompts on GitHub

Example prompts for breakup emails

Ask any of these in Claude, ChatGPT or any AI tool connected to Calven.

Write the breakup email for [persona] at [account] after [five] attempts.
What did buyers who said "not now" tell us once they bought?
Does this breakup read as honest to [persona]? [paste the email]
Work out why [persona] went silent using push, pull, anxiety and habit, then aim the breakup at the force that's winning.
Tell me how many sends I need before a breakup A/B test means anything, and design one that can actually conclude.
Give me a 60-word breakup email for [persona] on [pain].
Which won deal started with months of silence?
What yes-or-no question should I end on for [persona]?
Which objection do [persona]s raise most often after going quiet?
What do customers say made switching feel safe? Quote two.

Open the full prompts on GitHub

No new tool to learn

Calven works where your team already does: in Slack, and in Claude, ChatGPT or any AI tool through Calven MCP. Nobody opens another app to get the approved story.

In Slack

Mention @Calven in a thread or send it a DM. It answers from the approved version, with sources.

Leo Park9:12 AM

Have we worked Parcelwise before? Who did we talk to, and how did it end?

CalvenAgent9:12 AM

Yes. A deal closed lost in February at the proposal stage. The buyer said budget, not fit.

Who: Jonas Albrecht, Head of Data, was the champion and is still listed. Their CFO joined late and never saw a business case.

Open with: what changed since February, and Kestrel Payments, live in six weeks.

Sources: Parcelwise deal · CRM contacts · Win/loss survey

Example workspace. Names and numbers are illustrative.

In your AI tool

Add Calven to the AI workflows you already run. Every step reads from the approved version, with sources.

Question: Review this sequence for Heads of Data at mid-market fintechs before it goes live Monday.

Head of Data sequence, 4 touchesGoogle Doc

Used Calven MCP

Touch 1: opens on "observability", a word no customer used on calls. Rewritten to lead with "finance finds the broken report first."

Touch 3: "half Northwind's price" is out of date. Northwind cut its mid-tier price 15%. Swapped for the Kestrel Payments proof.

Missing: nothing answers "we already use Northwind", the brush-off Heads of Data gave on 6 of your last 10 calls. Touch 4 now does.

Head of Data sequence, revised.docxWord document

Here's the revised sequence, with every change marked.

Sources: Head of Data persona · Northwind battle card · Kestrel win story · Your last 10 calls

Example workspace. Names and numbers are illustrative.

What changes when BDR and SDR teams' AI tools read from Calven?

BDRs using AI without a shared source send fast emails every other vendor could have sent. Connected to Calven, every email, call and sequence starts from the same approved go-to-market context.

Pain point	Without Calven	With Calven
Research	Ten tabs of account research before every first touch	One question: fit, who to contact and why similar accounts bought
Openers	Generic openers about the product	Openers built on the pains buyers name on calls, in their words
Reply rates	Send and hope the buyer replies	Every draft tested on the persona before it goes out
Brush-offs	"We already use Northwind" ends the thread	Answers from buyers who switched from that competitor, with the quote
Accuracy	Proof lines and price claims copied from an old deck	Every claim checked against the approved product brief and battle cards
Handoffs	Two lines and a calendar invite for the AE	A handoff note with fit, buying role, pain and competitor filled in

Two weeks trial. You'll only need a day.

A day is all it takes to see what the agents can do for you.
You won't go back.

Start free trialNo credit card required

Questions, answered

How does Calven help SDRs write better cold emails?

Paste a draft into your AI tool and Calven reviews it as the buyer persona, built from your real customers. It tells you where the buyer stops reading and rewrites the email in the words customers use on calls, with a proof point from a won deal.

Can Calven tell BDRs which accounts to work?

Yes. Calven scores CRM accounts against the ICP the agents keep current and gives the reason for each score, so BDRs work the strong fits first and know which persona to contact.

How do BDRs handle "we already use a competitor" with Calven?

Ask how buyers you won from that competitor changed their minds. Calven answers from win/loss interviews and the battle card, with the quote and the deal behind it.

Do BDR and SDR teams need to log into Calven?

No. Product marketing runs Calven. Everyone else reaches the same approved context from the AI tool they already use over Calven MCP, or in Slack. Professional and Enterprise include unlimited seats, so each person signs in as themselves.

Which AI tools work with Calven?

Any client that supports remote MCP: Claude, ChatGPT, Cursor, VS Code with GitHub Copilot, Claude Code, Codex and others. Paste https://api.calven.ai/mcp, choose OAuth and sign in.