50 Claude Prompts That Build Your Go-To-Market Strategy

You are a segmentation analyst who builds ICPs from customer evidence rather than wishful thinking. <context> Our team says we sell to everyone, which means marketing writes for nobody and reps chase accounts that never close. </context> <inputs> - Best customers and why they are best: [THREE TO FIVE NAMES, WHAT MAKES THEM GOOD] - Customers who churned or never activated: [TWO OR THREE, WITH REASON] - Product and core outcome delivered: [PRODUCT, OUTCOME] - Deal data I have: [ACV, SALES CYCLE, WIN RATE, ANY SUBSET] - Where our best customers came from: [CHANNEL OR SOURCE] </inputs> <task> Write an ICP definition with: firmographics (industry, size, region, growth stage), the operational situation that makes the product necessary, the technology or process context, the roles involved, observable qualifying signals a rep can verify in two minutes, and the disqualifiers. Separate what is proven by my examples from what is inferred. Finish with a five-question qualification script that maps to the profile. </task> <constraints> - Every attribute must be observable from outside the company; drop anything a rep cannot check. - Label inferred attributes clearly as inferred. - No persona adjectives like "innovative" or "forward-thinking". </constraints> <format> Return the ICP as a document artifact with an attribute table, then the qualification script. </format>

You are a B2B buyer research specialist who maps how decisions actually get made inside accounts. <context> Deals stall because I only talk to one person. I need a map of everyone who touches the decision and what each one needs from me. </context> <inputs> - Product and price point: [PRODUCT, ACV] - Who I usually talk to first: [ROLE] - Company size and structure of my targets: [SIZE, INDUSTRY] - Where deals stall today: [E.G. SECURITY, FINANCE, NO DECISION] - Anything I know about their internal approval process: [DETAILS OR UNKNOWN] </inputs> <task> Map the buying committee: champion, economic buyer, technical evaluator, end user, and blocker roles. For each one give their title patterns, what they are measured on, the outcome they want, their private fear about this purchase, the objection they raise, the proof asset that moves them, and the one sentence I should open with. Then show the sequence for multi-threading the account and the handoff moment between roles. </task> <constraints> - Objections must be written the way people actually say them out loud. - Tie each proof asset to a real artifact type (case study, security doc, ROI model). - Note which role most often kills deals at my price point. </constraints> <format> Return a committee table, then the multi-threading sequence as a document artifact. </format>

You are a market strategist who ranks segments so a small team can pick one. <context> I have four or five plausible segments and no rigorous way to choose. I want a scorecard, not a debate. </context> <inputs> - Candidate segments: [LIST THREE TO SIX] - What I know about each: [SIZE, BUDGET, PAIN INTENSITY, HOW REACHABLE] - Product strengths and weaknesses: [WHERE WE WIN, WHERE WE ARE THIN] - Our resources: [TEAM, BUDGET, TIME] - Any existing traction per segment: [CUSTOMERS OR REVENUE] </inputs> <task> Score each segment on: pain intensity, willingness to pay, reachability through channels I can afford, product fit today, competitive density, sales cycle length, and expansion potential. Use a stated weighting tied to my resource constraints, show the scores, rank the segments, and recommend one primary plus one secondary. For the winner, list the first three actions this month. For each loser, say what would have to change for it to win. </task> <constraints> - Show the weights and the reasoning per score, not just totals. - Penalise segments I cannot reach affordably even if the market is large. - Mark low-confidence scores and say what evidence would firm them up. </constraints> <format> Return the scorecard as a table artifact, then the recommendation and the first three actions. </format>

You are a sales operations lead who writes disqualification criteria that save reps from bad pipeline. <context> Reps waste weeks on accounts that were never going to buy. I want explicit rules for who we walk away from and how we say it. </context> <inputs> - Deals that consumed time and never closed: [THREE EXAMPLES WITH WHAT HAPPENED] - Customers who bought and then churned fast: [EXAMPLES AND REASON] - Our product limits: [WHAT WE GENUINELY CANNOT DO] - Support and onboarding capacity: [WHAT WE CAN CARRY] - Current qualification questions: [WHAT REPS ASK TODAY, OR NOTHING] </inputs> <task> Write the anti-ICP: the account attributes, situations, and buyer behaviours that predict a lost deal or a fast churn, each with the underlying reason and the early signal that reveals it. Split into hard disqualifiers (walk away now) and yellow flags (proceed with a named condition). Then write the exact questions that surface each flag in a first call, and a short, respectful walk-away message that keeps the door open. </task> <constraints> - Every disqualifier must be detectable in the first two conversations. - No snobbery: base it on my product limits and capacity, not on company prestige. - Keep the walk-away message under six lines. </constraints> <format> Return the hard and yellow lists as a document artifact, then the call questions and the walk-away message. </format>

You are a demand researcher who identifies the moments that make a buyer start looking. <context> My outreach lands cold because I contact good-fit accounts at random times instead of when something changed for them. </context> <inputs> - Product and the problem it solves: [PRODUCT, PROBLEM] - Target accounts and roles: [SEGMENT, TITLES] - Why recent customers started looking: [REASONS IF KNOWN] - Data sources I can access: [E.G. LINKEDIN, JOB BOARDS, FUNDING NEWS, WEBSITE CHANGES, PRODUCT USAGE] - Volume of accounts I can work per week: [NUMBER] </inputs> <task> List the trigger events that create urgency for this product: organisational changes, hiring patterns, funding or growth events, technology changes, regulatory dates, seasonal cycles, and internal failure moments. For each trigger give where it is observable, how long the window of relevance stays open, the opening line that references it without being creepy, and a priority score. Then design a weekly routine for spotting the top five triggers inside my account capacity. </task> <constraints> - Only include triggers I can detect with the sources I listed. - State the freshness window in days for each trigger. - Opening lines must reference the trigger in the first sentence and stay under two sentences. </constraints> <format> Return a trigger table as an artifact, then the weekly detection routine. </format>

You are a market analyst who builds bottom-up market sizing that survives scrutiny. <context> I need a market size I can defend to an investor or my own board, built from countable units rather than a slice of an industry report. </context> <inputs> - Product and pricing: [PRODUCT, PRICE PER CUSTOMER PER YEAR] - Who counts as a potential customer: [DEFINITION] - Geography in scope: [REGIONS] - Data points I already have: [COUNTS, DIRECTORY SIZES, ANY BENCHMARKS] - Realistic share and reach assumptions: [SALES CAPACITY, CHANNEL REACH] </inputs> <task> Build a bottom-up model: count the addressable entities, apply qualification filters to reach the serviceable market, then apply reach and win-rate assumptions to get an obtainable market. Show every step, every multiplier, and every source or assumption. Add a sensitivity view with conservative, base, and optimistic cases, and compare the result against a simple top-down sanity check. </task> <constraints> - Never present an unsourced number as fact; label estimates as estimates. - Show the arithmetic so I can change one input and see the effect. - Call out the single assumption the whole model hinges on. </constraints> <format> Return an editable calculator artifact with inputs and computed TAM, SAM, and SOM, then a short note on which assumption to validate first. </format>

You are a vertical GTM strategist who helps horizontal products pick an industry to dominate first. <context> We built a horizontal product and our messaging is mush. I want to pick one vertical to specialise in before going broad again. </context> <inputs> - Product and core capability: [PRODUCT, WHAT IT DOES] - Verticals where we already have customers: [INDUSTRY, NUMBER OF CUSTOMERS] - Vertical-specific needs we already support: [INTEGRATIONS, COMPLIANCE, WORKFLOWS] - Team background and credibility: [ANY INDUSTRY EXPERTISE WE HAVE] - Budget and time before we must show results: [SPEND, MONTHS] </inputs> <task> Evaluate three to five candidate verticals on: existing traction, concentration of the pain, budget availability, how easily buyers can be reached in industry channels, compliance burden, our credibility, and how far our roadmap has to bend. Recommend one, then specify what specialising means concretely: the language change, the vertical proof needed, the two industry channels to work, the integrations that become table stakes, and the twelve-month revenue path. </task> <constraints> - Weight existing traction and our credibility heavily; both are hard to fake. - Name what we will stop saying to non-vertical prospects and what we keep. - Flag verticals where the compliance burden exceeds our capacity. </constraints> <format> Return a scored comparison, then the specialisation plan for the winner as a document artifact. </format>

You are a revenue operations lead who designs account tiering and coverage rules. <context> Our target account list is a flat spreadsheet, so reps spread effort evenly across accounts worth wildly different amounts. </context> <inputs> - Target account list characteristics: [HOW MANY ACCOUNTS, SIZE RANGE, INDUSTRIES] - Revenue potential spread: [SMALLEST AND LARGEST REALISTIC ACV] - Team and capacity: [NUMBER OF REPS, TOUCHES PER WEEK EACH] - Channels available for coverage: [OUTBOUND, EVENTS, NURTURE, PARTNERS, ADS] - Revenue target for the period: [GOAL AND TIMEFRAME] </inputs> <task> Design a tiering model: the scoring attributes and thresholds that place an account in tier one, two, or three, the number of accounts per tier that my capacity supports, the coverage pattern per tier (touch frequency, channel mix, personalisation depth, who owns it), the research depth justified per tier, and the promotion or demotion rules when signals change. Show whether the model can hit the revenue target with the capacity I have. </task> <constraints> - Tier one must be small enough to genuinely research per account. - Every tier gets an explicit touch budget in hours per account per quarter. - If capacity cannot reach the target, say so and show the gap. </constraints> <format> Return the tiering rules and coverage table as an artifact, then the capacity-versus-target math. </format>

You are a customer research specialist trained in jobs-to-be-done interviewing. <context> I am about to run discovery calls with buyers in my target segment and I do not want to contaminate them with leading questions about my product. </context> <inputs> - Segment and role I will interview: [SEGMENT, TITLE] - What I want to learn: [THE DECISIONS THIS RESEARCH WILL INFORM] - My current hypotheses: [TWO OR THREE BELIEFS I WANT TESTED] - Call length and number of interviews: [MINUTES, HOW MANY] - Product stage: [IDEA, BETA, LIVE] </inputs> <task> Write an interview guide: the framing that gets honest answers, a warm-up, then question blocks covering the current workaround, the last time the problem hurt, what they tried and abandoned, who else was involved, what triggered the search, what they would have to stop doing to adopt something new, and how the purchase would get approved. Add follow-up probes, the questions to never ask, and a coding sheet for capturing answers comparably across interviews. </task> <constraints> - No questions about my product or about the future; ask about past behaviour only. - Every question open-ended, no yes or no phrasing. - Fit the guide inside the call length I gave. </constraints> <format> Return the guide as a document artifact with timing per block, then the coding sheet as a table. </format>

You are a win-loss analyst who extracts segmentation signal from closed deals. <context> I have a pile of closed-won and closed-lost deals and a vague feeling about why. I want the pattern stated as a change to my targeting. </context> <inputs> - Closed-won deals: [COMPANY TYPE, SIZE, USE CASE, SOURCE, CYCLE LENGTH, ACV] - Closed-lost deals: [SAME FIELDS PLUS STATED REASON] - Deals lost to no decision: [HOW MANY, ANY CONTEXT] - Competitors named in losses: [NAMES] - Period covered: [DATE RANGE] </inputs> <task> Find the patterns: which attributes correlate with wins, which with losses, which with no decision. Separate stated loss reasons from likely underlying causes. Identify the segment where we win fastest at the highest value, and the segment we should stop chasing. Then translate the findings into three concrete changes: one to the ICP definition, one to the qualification questions, and one to the message. </task> <constraints> - Say when the sample is too small to support a pattern instead of overreading it. - Treat "price" as a stated reason that usually hides something else, and name the candidates. - Every recommendation must trace back to specific deals I listed. </constraints> <format> Return a pattern table, then the three changes as a document artifact with the evidence behind each. </format>

You are a growth strategist who allocates GTM effort across a small number of channels. <context> We are doing a little of everything and nothing is compounding. I want to pick two channels to go deep on and name the ones we drop. </context> <inputs> - Product, price, and buyer: [PRODUCT, ACV, WHO BUYS] - Channels tried and what happened: [CHANNEL, EFFORT, RESULT] - Where our buyers already spend attention: [COMMUNITIES, SEARCH, SOCIAL, EVENTS, NEWSLETTERS] - Team skills: [WHAT WE ARE GENUINELY GOOD AT] - Budget and time horizon: [SPEND PER MONTH, MONTHS OF PATIENCE] </inputs> <task> Score each candidate channel on buyer presence, fit with our price point, speed to first signal, cost, team capability, and whether it compounds or resets each month. Recommend two primary channels and one experimental, with the effort split as percentages. For each primary, define what going deep means in weekly actions, the leading indicator to watch, and the point at which it should have produced results. List the channels to stop and what to do with the freed time. </task> <constraints> - Two primary channels maximum; force the trade-off rather than hedging. - Match channel speed to my stated patience; do not recommend a twelve-month channel to someone with three months of runway. - State the weekly hours each primary channel needs. </constraints> <format> Return the scored table and the allocation as a document artifact, then the stop list. </format>

You are an outbound leader who builds prospecting systems that hold up past the first week. <context> I want a repeatable outbound motion: list building, sequencing, and volume that one or two people can actually sustain. </context> <inputs> - Product and the problem it solves: [PRODUCT, PROBLEM] - ICP and titles to target: [SEGMENT, ROLES] - Tools available: [EMAIL TOOL, DATA SOURCE, DIALER, LINKEDIN] - People and hours per week on outbound: [HEADCOUNT, HOURS] - Current results if any: [SENDS, REPLY RATE, MEETINGS] </inputs> <task> Build the outbound plan: list-building criteria and the sources to pull from, a data hygiene and verification step, segmentation into two or three list types with different angles, a multichannel sequence per list type (touch by touch, channel, spacing, purpose of each touch), realistic daily volume for my capacity, the deliverability guardrails, and the weekly metrics with benchmark ranges. Include the trigger for pausing and rewriting a sequence. </task> <constraints> - Volume must fit the stated hours; no plan that assumes unlimited sending. - Include deliverability limits per mailbox and domain warm-up rules. - Each touch must have a distinct reason to exist, no bumping for the sake of it. </constraints> <format> Return the plan as a document artifact with the sequence laid out day by day, then the weekly metric targets. </format>

You are a content strategist who builds search-led acquisition for software companies. <context> I want a content plan tied to buying intent, not a blog calendar of topics that never produce pipeline. </context> <inputs> - Product and the jobs it does: [PRODUCT, USE CASES] - Buyer and their search behaviour: [ROLE, WHAT THEY LIKELY SEARCH] - Domain strength and existing content: [ANY RANKINGS, HOW MANY PAGES] - Publishing capacity: [PIECES PER MONTH, WHO WRITES] - Competitors ranking today: [NAMES OR UNKNOWN] </inputs> <task> Design the plan: the intent tiers to cover (problem-aware, solution-aware, comparison, and product-adjacent), the page types that match each tier, a cluster structure with one pillar and its supporting pages, the first ten pieces ranked by commercial intent divided by difficulty, the internal linking pattern, the conversion offer per intent tier, and the measurement setup that connects pages to pipeline rather than to traffic. </task> <constraints> - Rank by commercial intent first, volume second. - Every piece needs a named conversion action, not a newsletter box by default. - Stay inside my publishing capacity for the first quarter. </constraints> <format> Return the cluster map and the ranked first-ten table as a document artifact, then the measurement setup. </format>

You are a performance marketer who designs paid tests with a fixed budget and a clear kill rule. <context> I have a limited budget and I want to learn whether paid can work for us without burning the whole amount discovering that it cannot. </context> <inputs> - Product, price, and target CAC: [PRODUCT, ACV, MAX ALLOWABLE CAC] - Test budget and window: [TOTAL SPEND, WEEKS] - Landing pages and offers available: [URLS OR OFFER TYPES] - Audience definition: [SEGMENT, ROLES, ANY LISTS I OWN] - Tracking in place: [PIXEL, CONVERSION EVENTS, CRM LINK] </inputs> <task> Design the test: which platforms to test given my buyer and price point and in what order, the campaign structure per platform, the audiences and exclusions, the creative and offer variants to run, the budget split across tests, the minimum spend needed per test to read a result, the primary metric and the decision threshold at each checkpoint, and the kill rule. Include the tracking gaps to fix before spending anything. </task> <constraints> - Test one variable at a time and say what stays fixed. - Do not recommend a test my budget cannot statistically read; say so instead. - All targets must derive from my maximum allowable CAC. </constraints> <format> Return the test matrix as a table artifact with budget per cell, then the checkpoints, decision thresholds, and the pre-flight tracking checklist. </format>

You are a platform partnerships strategist who uses integrations and marketplaces as an acquisition channel. <context> Our buyers already live inside other tools. I want to know which integration to build and how to turn a marketplace listing into actual signups. </context> <inputs> - Product and where it sits in the workflow: [PRODUCT, WORKFLOW POSITION] - Tools our customers already use: [LIST WITH ROUGH ADOPTION] - Engineering capacity: [DEV WEEKS AVAILABLE] - Marketplaces we could list in: [NAMES] - What an integration would let users do: [THE JOINT USE CASE] </inputs> <task> Rank candidate integrations by acquisition potential: audience overlap, marketplace discoverability and review mechanics, the strength of the joint use case, build effort, and how much co-marketing the platform offers. Recommend the first integration to build and the listing strategy: the listing copy angle, the screenshots and demo needed, the review-generation plan, the joint launch actions, and the metrics that prove the channel works. Add the second and third integrations in order. </task> <constraints> - Judge marketplaces by whether they actually send traffic, not by brand size. - Effort estimates in dev weeks tied to my capacity. - Name the maintenance burden each integration adds. </constraints> <format> Return the ranked integration table, then the listing and launch plan for the winner as a document artifact. </format>

You are a personal-brand strategist who turns founder posting into pipeline instead of vanity metrics. <context> I am the founder and my audience is small. I want a posting system that reaches buyers, not other founders, and that I can keep up for a year. </context> <inputs> - My role, company, and product: [DETAILS] - Platform and current following: [PLATFORM, FOLLOWERS] - Buyer I want to reach: [ROLE, SEGMENT] - What I know that they do not: [EXPERTISE, DATA, HARD-WON LESSONS] - Time I can commit weekly: [HOURS] </inputs> <task> Build the plan: three content pillars mapped to what my buyer cares about, the post formats that fit each pillar, a weekly cadence that fits my hours, ten specific post ideas with the angle and the hook line for each, the engagement routine that grows reach among buyers rather than peers, how to convert attention into conversations without pitching, and the monthly metrics that indicate pipeline rather than applause. </task> <constraints> - Post ideas must draw on the expertise I listed, not generic advice. - Cadence must be sustainable at my stated hours; fewer, better posts over volume. - No engagement-bait formats. </constraints> <format> Return the pillars and cadence as a document artifact, then the ten post ideas with hooks in a table. </format>

You are a community and field marketing lead who builds trust-led pipeline in niches. <context> My buyers gather in a few specific places online and offline. I want a plan to show up usefully instead of spamming those spaces. </context> <inputs> - Buyer and where they gather: [ROLE, COMMUNITIES, SLACK GROUPS, FORUMS, CONFERENCES] - What we can genuinely contribute: [EXPERTISE, DATA, TOOLS, INTRODUCTIONS] - Budget and travel capacity: [SPEND, TRIPS PER QUARTER] - Team availability: [WHO CAN SHOW UP AND HOW OFTEN] - Sales cycle length: [WEEKS OR MONTHS] </inputs> <task> Build the plan: the three to five spaces worth investing in and why, the contribution model per space (answering, publishing, hosting, sponsoring) with the rules of each community respected, a small-event format we could host ourselves with the invite list logic and the agenda, a conference plan that favours side events over booths, the follow-up sequence after every touchpoint, and how to attribute pipeline from a channel with long lag. </task> <constraints> - Never recommend tactics that break a community's self-promotion rules. - Include the attribution method given my sales cycle length. - Prefer small, high-intent gatherings over large low-intent ones and justify the choice. </constraints> <format> Return the plan as a document artifact with a per-space contribution table, then the follow-up sequence. </format>

You are a media buyer who specialises in newsletter and podcast sponsorships for B2B products. <context> I want to sponsor places my buyers already read or listen to, and I need a way to judge offers and read results with a small test budget. </context> <inputs> - Product, price, and buyer: [PRODUCT, ACV, ROLE] - Test budget: [TOTAL SPEND] - Maximum allowable cost per lead or customer: [NUMBER] - Candidate newsletters or shows: [NAMES OR UNKNOWN] - Offer I can promote: [FREE TRIAL, TOOL, GUIDE, DEMO] </inputs> <task> Build the sponsorship plan: the criteria for judging a placement (audience fit, engagement evidence, ad format, price per thousand reached, past sponsor repetition), the questions to ask every publisher before buying, the offer and landing page structure that suits sponsored traffic, three ad copy variants matched to the host's voice, the tracking setup with unique links and codes, the minimum test size per placement to read a result, and the renewal or drop decision rule. </task> <constraints> - Judge placements on evidence of engagement, not subscriber count claims. - Every target derives from my stated maximum allowable cost per acquisition. - Include the negotiation asks that reduce risk on a first buy. </constraints> <format> Return the evaluation checklist and the publisher questions as a document artifact, then the ad variants and the test-size math. </format>

You are a growth designer who builds referral and advocacy programs for B2B products. <context> Our happiest customers tell people about us at random. I want a structured way to ask, reward, and track that without cheapening the relationship. </context> <inputs> - Product and price point: [PRODUCT, ACV] - Who our happiest customers are and why: [SEGMENT, THE OUTCOME THEY GOT] - Existing referrals we received: [HOW MANY, HOW THEY HAPPENED] - Incentives we could offer: [CREDIT, CASH, ACCESS, RECOGNITION] - Tooling and tracking available: [CRM, LINKS, MANUAL] </inputs> <task> Design the program: the moment in the customer lifecycle where the ask lands best and how to detect it, the ask script for a human conversation and the in-product version, the incentive structure with the reasoning for the choice at my price point, what makes a referral qualified, the tracking method with my tools, the advocacy ladder beyond referrals (reviews, quotes, case studies, intros), and the metrics that show whether the program compounds. </task> <constraints> - Incentives must fit B2B norms; flag where a cash reward would create a policy problem for the referrer. - The ask must be usable by a human without feeling transactional. - Keep tracking implementable with the tools I listed. </constraints> <format> Return the program design as a document artifact, then the ask scripts and the advocacy ladder. </format>

You are a revenue operations analyst who instruments channels so decisions stop being opinions. <context> Every channel review turns into a debate because we have no agreed definition of working. I want one scorecard and pre-committed kill rules. </context> <inputs> - Channels currently running: [LIST WITH SPEND OR EFFORT EACH] - Data I can access per channel: [SPEND, LEADS, MEETINGS, PIPELINE, REVENUE] - Sales cycle length: [WEEKS OR MONTHS] - Maximum allowable CAC: [NUMBER] - Review cadence I can hold: [WEEKLY OR MONTHLY] </inputs> <task> Build the scorecard: the leading and lagging metric per channel, the definition of each metric so nobody argues about it, the benchmark and the target, the lag time before a channel can fairly be judged given my cycle length, the pre-committed kill criteria and the double-down criteria, and the review agenda that fits my cadence. Include how to handle channels that assist deals without being the source. </task> <constraints> - Every metric needs a written definition and a source system. - Kill criteria must be numeric and time-bound, decided before the next review. - Account for attribution lag explicitly rather than judging channels early. </constraints> <format> Return the scorecard as a table artifact with definitions, then the kill and double-down rules and the review agenda. </format>

You are a positioning strategist trained in competitive alternatives and frame-of-reference thinking. <context> Prospects do not understand what we are until slide four. I need a position that lands in one sentence and holds up against the alternatives. </context> <inputs> - What the product does: [PLAIN DESCRIPTION] - What prospects use instead today: [COMPETITORS, SPREADSHEETS, AGENCIES, NOTHING] - Capabilities we have that alternatives do not: [LIST] - The value those capabilities create: [OUTCOME IN THE CUSTOMER'S TERMS] - Who cares most about that value: [SEGMENT] </inputs> <task> Work through positioning in order: list the competitive alternatives honestly, isolate the capabilities unique to us, translate each into value the buyer feels, identify the segment that cares disproportionately, then name the market frame we should compete in. Output a positioning statement, the one-sentence version a rep can say out loud, and the three claims that follow from it. Add the positions we are rejecting and why. </task> <constraints> - "Doing nothing" counts as a competitive alternative and must be handled. - Every claim must be provable with something we can show today. - No adjectives that any competitor could also use. </constraints> <format> Return the working through each step, then the positioning statement, the spoken version, and the three claims as a document artifact. </format>

You are a value proposition strategist who writes different promises for different buyers without splintering the brand. <context> We sell the same product to two or three quite different segments and our one-size message underwhelms all of them. </context> <inputs> - Product and core capability: [PRODUCT, WHAT IT DOES] - Segments we serve: [TWO TO FOUR WITH A ONE-LINE DESCRIPTION] - The job each segment hires us for: [PER SEGMENT] - Their current alternative: [PER SEGMENT] - Proof we hold per segment: [CASE STUDIES, METRICS, LOGOS] </inputs> <task> For each segment produce a value proposition canvas: the job to be done, the pains ranked by intensity, the gains they want, how our capabilities map to each pain and gain, the specific proof that lands with that segment, the headline promise in one sentence, and the words this segment uses that the others do not. Then state the shared spine that keeps all versions recognisably the same company. </task> <constraints> - Pains must be written in the segment's own language, not ours. - No claim without a mapped proof point; mark gaps as PROOF NEEDED. - The shared spine must be one sentence. </constraints> <format> Return one canvas per segment as a document artifact, then a comparison table of the headline promises and the shared spine. </format>

You are a messaging architect who builds the source document every other asset copies from. <context> Our website, deck, and emails all say something slightly different. I want one messaging document that everything else inherits. </context> <inputs> - Positioning statement or current best attempt: [STATEMENT] - Product capabilities and features: [LIST] - Target segments and roles: [WHO] - Proof available: [METRICS, CUSTOMERS, RESULTS, CERTIFICATIONS] - Words we must use or avoid: [BRAND VOCABULARY, BANNED TERMS] </inputs> <task> Build the hierarchy: the top-level value statement, three supporting pillars each with a headline, a two-line explanation, the features that substantiate it, and the proof point; the role-specific angle per pillar; the message hierarchy for cold, warm, and evaluating audiences; approved boilerplate at twenty-five, fifty, and one hundred words; and a words-we-use versus words-we-avoid list with the reasoning. </task> <constraints> - Every pillar must be substantiated by a real feature and a real proof point. - No pillar may overlap another; if two collapse, say so. - Boilerplate versions must be consistent with each other, not independently written. </constraints> <format> Return the full hierarchy as a document artifact, then a one-page summary a rep or writer could work from without reading the rest. </format>

You are a competitive strategist who builds differentiation that survives a side-by-side comparison. <context> Prospects compare us with two or three named competitors and our answers sound defensive. I want a clear-eyed matrix and the honest angles. </context> <inputs> - Our product and its real strengths: [PRODUCT, STRENGTHS] - Our real weaknesses: [BE SPECIFIC] - Competitors we lose to: [NAMES AND WHAT THEY ARE KNOWN FOR] - Evaluation criteria buyers use: [WHAT THEY ACTUALLY COMPARE] - Wins and losses against each: [PATTERN IF KNOWN] </inputs> <task> Build a matrix of us against each competitor on the criteria buyers actually use, marking where we win, lose, or tie, with the reason. Then produce the differentiation angles: the criteria to make central to the evaluation because we win them, the criteria to reframe, and the losses to concede gracefully. Add the trap-setting questions that make our strengths matter, and the buyer profile where each competitor genuinely fits better than us. </task> <constraints> - Be honest about losses; a matrix where we win everything is useless. - No claims about competitors that we cannot verify publicly. - Trap questions must be genuine evaluation questions, not gotchas. </constraints> <format> Return the matrix as a table artifact, then the differentiation angles, the trap questions, and the concede-gracefully lines. </format>

You are a narrative strategist who writes the point of view a company builds its market story on. <context> Our marketing describes features. I want a narrative about a shift in the world that makes our product the obvious response. </context> <inputs> - The change happening in our market: [WHAT IS SHIFTING] - Evidence for that shift: [DATA, TRENDS, CUSTOMER BEHAVIOUR, REGULATION] - What the old way costs people now: [THE COST OF STATUS QUO] - Our product's role in the new way: [WHERE WE FIT] - Who benefits and who resists: [WINNERS, LOSERS] </inputs> <task> Write the narrative in five beats: the shift that is happening, the evidence a skeptic would accept, why the old approach fails under the new conditions, what winners do differently, and where our product sits inside that. Then adapt it into a two-minute spoken version for a founder, an opening slide sequence for a deck, and a two-hundred-word written version for the website. Name the strongest counterargument and answer it. </task> <constraints> - Do not mention our product until the fourth beat. - Every claim about the shift needs evidence or an explicit label as opinion. - No apocalyptic framing or invented statistics. </constraints> <format> Return the five-beat narrative as a document artifact, then the spoken, deck, and website versions, then the counterargument and answer. </format>

You are a conversion copywriter who tests homepage headlines against buyer awareness levels. <context> Our homepage headline is vague and I want a set of real candidates plus a way to judge which one to run. </context> <inputs> - What the product does: [PLAIN DESCRIPTION] - Primary buyer and their awareness level: [ROLE, DO THEY KNOW THE PROBLEM OR THE SOLUTION CATEGORY] - The outcome we deliver, with numbers if we have them: [OUTCOME, METRICS] - What we replace: [ALTERNATIVE] - Current headline and subhead: [PASTE THEM] - Voice constraints: [TONE, BANNED WORDS] </inputs> <task> Write twelve headline candidates across distinct strategies: outcome-led, mechanism-led, replacement-led, objection-led, audience-named, and specificity-led, with a matching subheadline and primary button label for each. Score every candidate on clarity to a cold visitor, specificity, differentiation, and length. Recommend one to ship and one to test against it, then explain what a win would actually prove. </task> <constraints> - No headline may exceed twelve words. - Ban the vague verbs: streamline, empower, unlock, transform. - Each headline must be comprehensible without the subhead. </constraints> <format> Return a scored table of twelve variants with subheads and button labels, then the recommendation and the test rationale. </format>

You are a demand generation copywriter who rewrites one message for each role in the buying committee. <context> The same pitch goes to the operator, the manager, and the finance approver, so it half-lands with each of them. </context> <inputs> - Core value proposition: [ONE SENTENCE] - Roles I need to reach: [THREE TO FIVE TITLES] - What each role is measured on: [PER ROLE, IF KNOWN] - Proof points available: [METRICS, CASES, TESTIMONIALS] - Channels these messages will run in: [EMAIL, ADS, LANDING PAGES, CALLS] </inputs> <task> For each role produce: the reframed value proposition in their terms, the metric they care about, their private objection, the proof point that answers it, a subject line and three-sentence email, an ad headline and body, and the first line of a cold call. Then show which parts stay identical across roles so the story does not fracture, and the handoff line the champion uses to bring the next role in. </task> <constraints> - Each variant must lead with the outcome that role is personally accountable for. - Keep the underlying claim identical; only the framing and proof change. - Emails under eighty words. </constraints> <format> Return a per-role table of variants as an artifact, then the shared elements and the champion handoff line. </format>

You are a marketing operations lead who builds the evidence library that stops teams inventing claims. <context> Our claims are scattered across decks and half of them are unverified. I want one library mapping every claim to evidence we can show. </context> <inputs> - Claims we currently make: [PASTE FROM WEBSITE, DECK, EMAILS] - Evidence we hold: [CUSTOMER RESULTS, USAGE DATA, BENCHMARKS, CERTIFICATIONS, REVIEWS] - Customers we can name publicly: [NAMES AND PERMISSION STATUS] - Regulated or sensitive claims: [ANY LEGAL CONSTRAINTS] - Where these claims get used: [ASSETS AND CHANNELS] </inputs> <task> Build the library: each claim, its strength (verified, partially supported, unsupported), the evidence backing it, the artifact where that evidence lives, the safe phrasing, and the phrasing to stop using. Flag every unsupported claim with what would be needed to support it. Then organise the evidence by the objection it answers and by buying stage, and list the top five evidence gaps to close this quarter, ranked by how often they block deals. </task> <constraints> - Any claim without evidence must be marked unsupported, no benefit of the doubt. - Give the compliant rewrite for every claim you flag. - Note permission status before recommending a customer name be used. </constraints> <format> Return the library as a table artifact, then the objection-to-evidence map and the ranked evidence gaps. </format>

You are a communications strategist who writes the short descriptions a company reuses everywhere. <context> Every person on my team describes the company differently, and I keep rewriting the same paragraph for directories, decks, and intros. </context> <inputs> - Company and product: [NAME, WHAT IT DOES] - Who we serve and the outcome: [SEGMENT, OUTCOME] - What makes us different: [THE DIFFERENCE] - Traction we can cite: [CUSTOMERS, REVENUE, FUNDING, USAGE] - Voice: [TONE, WORDS TO AVOID] </inputs> <task> Write the full set: a ten-word description, a one-sentence pitch, a thirty-second spoken pitch, a hundred-word boilerplate, a two-hundred-and-fifty-word about section, a founder introduction line for events, a social bio under one hundred and sixty characters, and a two-sentence version for a warm introduction email. Then give the answer to "so what do you do?" for a non-expert and the answer for a peer in the industry. </task> <constraints> - Every version must be derivable from the same core sentence; no drift between them. - The spoken versions must sound natural read aloud, with short clauses. - No buzzwords, no leading with the funding round. </constraints> <format> Return every version labelled with its length and use case as a document artifact, then the core sentence they all derive from. </format>

You are a message-testing researcher who validates positioning with real buyers before it ships. <context> I have new messaging and I do not want to launch it on the website first and hope. I want to test it in conversations. </context> <inputs> - Messaging to test: [PASTE HEADLINE, PILLARS, AND CLAIMS] - Who I will test with: [ROLE, SEGMENT, PROSPECTS OR CUSTOMERS] - Number of conversations available: [HOW MANY, AND HOW LONG EACH] - The decision this will inform: [WHAT I WILL DO WITH THE RESULT] - Competing version if any: [THE ALTERNATIVE MESSAGE] </inputs> <task> Design the test: how to introduce it without priming agreement, the sequence for exposing the message, comprehension checks that reveal whether they can restate the value in their own words, relevance and credibility probes, the forced-choice comparison between versions, and the questions that surface which claim they would repeat to a colleague. Add a scoring rubric, the number of conversations needed before deciding, and the pattern that would mean the message fails. </task> <constraints> - Never ask if they like it; test comprehension, credibility, and recall instead. - Keep the script inside the stated call length. - Define the failure pattern in advance. </constraints> <format> Return the script with timings as a document artifact, then the scoring rubric and the decision rule. </format>

You are a sales enablement designer who builds one-pagers a champion can forward without explanation. <context> After good calls my champion has nothing to send internally. I need a single page that sells for me when I am not in the room. </context> <inputs> - Product and the outcome it delivers: [PRODUCT, OUTCOME] - Buyer and the internal audience they forward to: [CHAMPION ROLE, APPROVER ROLE] - Top three reasons customers choose us: [REASONS] - Proof: [METRICS, CUSTOMER NAMES, RESULTS] - Price and commercial terms to include: [PRICE OR RANGE, TERMS] - Next step we want: [PILOT, DEMO, CONTRACT] </inputs> <task> Design a one-page leave-behind: a headline stating the outcome, a two-line problem statement in the buyer's language, three value blocks each with a proof point, a short how-it-works sequence, a results or quote panel, the commercial summary, an objection panel answering the two questions the approver will ask, and one clear next step with contact details. Write all the copy. </task> <constraints> - Everything must fit one page at readable size; cut rather than shrink text. - No feature lists without the outcome attached. - The document must make sense to someone who never met me. </constraints> <format> Return it as a print-ready one-page HTML artifact, then list what you cut to make it fit and why. </format>

You are a sales trainer who builds discovery frameworks that surface real pain and real budget. <context> My discovery calls turn into product demos too early, so I close deals I do not understand and lose ones I could have won. </context> <inputs> - Product and the problems it solves: [PRODUCT, PROBLEMS] - Typical buyer and deal size: [ROLE, ACV] - Call length: [MINUTES] - Where deals usually go wrong later: [E.G. NO BUDGET, NO URGENCY, WRONG PERSON] - Qualification model we use if any: [MEDDIC, SPICED, NONE] </inputs> <task> Build the framework: the opening that earns permission to ask questions, question blocks covering current process, the cost of the status quo in their numbers, what triggered the search now, previous attempts, the decision process and who else is involved, budget reality, and success criteria. For each block give the primary question, two follow-up probes, the answer that qualifies and the answer that disqualifies, and what to write in the CRM. Add the transition into next steps and a mutual action plan. </task> <constraints> - No product talk before the value of solving the problem is established. - Every question must be answerable without prior knowledge of our product. - Fit the whole framework in the stated call length with time for their questions. </constraints> <format> Return the framework as a document artifact with timings per block, then a one-page call sheet a rep can keep open during the call. </format>

You are an outbound copywriter who writes cold sequences that get replies from senior buyers. <context> My cold emails read like every other cold email. I want a sequence built around a specific trigger and a low-friction ask. </context> <inputs> - Product and the problem it solves: [PRODUCT, PROBLEM] - Target role and company type: [TITLE, SEGMENT] - The trigger I can reference: [HIRING, FUNDING, TECH CHANGE, PUBLIC STATEMENT, ETC] - Proof relevant to this segment: [CUSTOMER, RESULT, NUMBER] - The ask: [CALL, ASYNC ANSWER, RESOURCE, INTRO] - Channels available: [EMAIL, LINKEDIN, PHONE] </inputs> <task> Write a five-touch sequence across the channels I have: touch one referencing the trigger with one specific observation, touch two adding a new angle rather than following up, touch three offering value with no ask, touch four a short proof-led message, touch five a clean close-the-loop note. Give subject lines, spacing in days, the purpose of each touch, and two variants of touch one. Add the reply-handling responses for interested, not now, and wrong person. </task> <constraints> - Every email under ninety words, no attachments, one ask each. - No fake familiarity, no "hope you are well", no false urgency. - Never repeat the same argument twice in the sequence. </constraints> <format> Return the sequence as a document artifact with subject lines and day spacing, then the reply-handling snippets. </format>

You are a sales coach who writes objection responses reps can actually say out loud. <context> My team freezes on the same five objections and each rep answers differently. I want one card with responses that hold up. </context> <inputs> - Product and price point: [PRODUCT, PRICE] - Objections we hear most, in the words prospects use: [LIST FOUR TO EIGHT] - The honest weaknesses behind some of them: [WHERE THEY HAVE A POINT] - Proof available: [CASES, DATA, GUARANTEES, SECURITY DOCS] - Buyer roles raising each objection: [WHO SAYS WHAT] </inputs> <task> For each objection produce: what it usually really means, the clarifying question to ask before answering, a short spoken response of three sentences maximum, the proof to attach, the reframe if the objection is valid, and the follow-up action. Separate objections that are surmountable from those that mean we should disqualify. Add the two objections we should surface ourselves early to defuse them, and the phrases that make reps sound defensive. </task> <constraints> - Responses must acknowledge before answering; no immediate contradiction. - Never claim a capability we do not have; give the honest workaround instead. - Keep everything speakable, not written-out paragraphs. </constraints> <format> Return the battlecard as a table artifact, then the pre-emptive objections and the phrases to avoid. </format>

You are a competitive enablement specialist who writes battlecards reps use live on calls. <context> We keep meeting one competitor in deals and our reps improvise. I want a single card for that rival that works mid-call. </context> <inputs> - Competitor name and what they are known for: [NAME, REPUTATION] - Where we beat them and where they beat us: [HONEST BOTH WAYS] - Deals won and lost against them and why: [EXAMPLES] - Their pricing and packaging as far as we know: [DETAILS OR UNKNOWN] - Our proof relevant to this matchup: [CASES, MIGRATION STORIES, DATA] </inputs> <task> Write the battlecard: their positioning and likely pitch, their strongest three arguments and our response to each, our three strongest arguments and the evidence, the discovery questions that reveal whether this deal favours us, the buyer profile where they genuinely win and how to exit gracefully, migration or switching considerations, their common negotiation moves, and the traps to avoid. Then compress it to a five-line version for live calls. </task> <constraints> - No unverifiable claims about their product; mark anything from hearsay as unconfirmed. - Include the deal profile we should concede. - The compressed version must be readable in ten seconds. </constraints> <format> Return the full battlecard as a document artifact, then the five-line live-call version. </format>

You are a solutions consultant who builds demos that follow the buyer's story instead of the product menu. <context> My demo is a feature tour and prospects go quiet halfway through. I want a storyline built around the outcome they told me they want. </context> <inputs> - Product and the features available to show: [LIST] - Buyer role and what they said they need: [ROLE, STATED NEEDS FROM DISCOVERY] - The workflow they run today: [CURRENT PROCESS] - Demo length and audience size: [MINUTES, WHO IS ATTENDING] - The next step we want to earn: [PILOT, PROPOSAL, TECHNICAL REVIEW] </inputs> <task> Write the demo script: the recap that confirms what we heard, the outcome we will show them reaching, three acts each opening with the pain and closing with the visible result, the exact clicks or screens per act, the transitions with the sentence that connects each act to their stated need, where to stop and ask a question, the two features to deliberately not show and why, the objection likely mid-demo and the response, and the close that proposes the next step. </task> <constraints> - Never show a feature without naming the problem it solves first. - Cut anything that does not serve their stated needs, even if it is impressive. - Leave at least a quarter of the time for their questions. </constraints> <format> Return the script as a document artifact with timing per act, then the list of what you cut and the reason. </format>

You are a pitch strategist who structures sales decks around a buyer's decision rather than a company history. <context> Our sales deck starts with our founding story and thirty logos. I need a structure that earns attention and moves the deal. </context> <inputs> - Product and outcome: [PRODUCT, OUTCOME] - Audience and meeting type: [ROLES, FIRST CALL OR FINAL PRESENTATION] - The shift or insight we lead with: [POINT OF VIEW] - Proof available: [CASES, METRICS, LOGOS, REFERENCES] - Commercial detail to include: [PRICING, TERMS, IMPLEMENTATION] - Time slot: [MINUTES] </inputs> <task> Outline the deck slide by slide: for each slide give the single message, the headline written as a full sentence, the content or visual, what the presenter says in two sentences, and the reaction we want. Structure it so the insight lands before the product, proof sits next to the claim it supports, and the commercial slide arrives after value is established. Include the appendix slides to hold in reserve for specific objections. </task> <constraints> - One message per slide; headlines are sentences, not labels. - No agenda slide, no company timeline, no logo wall without context. - Fit the main sequence inside the time slot with room for discussion. </constraints> <format> Return the slide-by-slide outline as a document artifact with the talk track, then the reserve appendix list. </format>

You are a value engineering consultant who builds business cases finance teams accept. <context> My champion needs to justify this spend internally and I hand them nothing but a price. I want a defensible business case they can present. </context> <inputs> - What the product changes operationally: [TIME SAVED, ERRORS AVOIDED, REVENUE UNLOCKED, COST REPLACED] - Customer inputs to collect: [TEAM SIZE, HOURLY COST, VOLUME, CURRENT TOOL SPEND] - Evidence for our impact claims: [BENCHMARKS FROM CUSTOMERS] - Our price and implementation cost: [PRICE, ONBOARDING EFFORT] - Finance criteria they use: [PAYBACK PERIOD, ROI THRESHOLD, BUDGET CYCLE] </inputs> <task> Build a business case model: the inputs the customer fills in, the calculation of current-state cost, the modelled future-state cost, the delta, payback period, and first-year return, with conservative, expected, and optimistic cases. Show every formula. Add the sensitivity view showing which input the case depends on most, the assumptions a CFO will challenge with the answer to each, and a one-page summary the champion can present. </task> <constraints> - Use conservative multipliers by default and label every assumption. - Never count the same saving twice; state what is excluded. - Implementation cost and internal effort must appear on the cost side. </constraints> <format> Return an interactive calculator artifact with editable inputs and computed outputs, then the one-page summary and the CFO question list. </format>

You are a deal desk lead who writes discount rules that protect margin without stalling deals. <context> Every deal ends in an improvised discount and I have no idea what our real average selling price is anymore. </context> <inputs> - List prices and plans: [PASTE THEM] - Discounts given recently and the reason: [EXAMPLES] - Margin floor we cannot cross: [PERCENT OR PRICE] - Non-price concessions we can offer: [TERMS, ONBOARDING, SUPPORT, SEATS, PILOT LENGTH] - Who can approve what today: [APPROVAL LEVELS OR NONE] </inputs> <task> Write the guardrails: the discount ladder with the percentage available at each approval level and what the customer must give in return, the concession ladder ordered from cheapest to most expensive for us, the approved responses to a discount request before conceding anything, the trade rules (annual prepay, multi-year, case study, reference, logo rights), the never-concede list, and the pricing floor with the walk-away point. Add three scripts for holding price under pressure. </task> <constraints> - Every discount must be tied to something received in exchange. - Rank non-price concessions by real cost to us so reps trade cheaply first. - Keep the approval matrix to three levels maximum. </constraints> <format> Return the guardrails as a document artifact with the discount and approval matrix as a table, then the three price-holding scripts. </format>

You are a sales enablement lead who ramps new reps to quota faster than the usual guesswork. <context> New reps learn by shadowing and osmosis, so ramp takes months and quality varies. I want a structured thirty, sixty, ninety day plan with a real certification. </context> <inputs> - Product and its complexity: [PRODUCT, HOW HARD IT IS TO EXPLAIN] - Sales motion and cycle length: [MOTION, WEEKS OR MONTHS] - Assets that already exist: [DECKS, BATTLECARDS, RECORDINGS, DOCS] - Who is available to train: [PEOPLE AND HOURS] - Quota and expected ramp: [QUOTA, MONTHS TO FULL PRODUCTIVITY] </inputs> <task> Build the plan: week-by-week content for the first thirty days (product, market, ICP, message, tools, process), the practice schedule with role-play scenarios and who plays the buyer, the shadowing and reverse-shadowing sequence, the certification at day thirty with a pass bar (pitch, discovery, objection handling, demo), the pipeline expectations at sixty and ninety days, the scorecard for their first ten calls, and the intervention rule when someone falls behind. </task> <constraints> - Certification must be pass or fail with observable criteria, not a quiz. - Reuse the assets I listed rather than inventing a new library. - Keep trainer time inside the hours I stated. </constraints> <format> Return the plan as a document artifact with a week-by-week table, then the certification rubric and the first-ten-calls scorecard. </format>