Prompt like a Marketer

Your AI Prompt Library

Authors:

Corrina Owens & Elaine Zelby

Why This Exists

Prompting is a new way of thinking.

And it's our belief that it's the single biggest skill that marketers can learn for current and future roles.

Marketers have accepted we’re in the era of AI, where virtually anything in our go-to-market motion can be improved with just a few prompts. Yet we’re continuously asking how we can improve at prompting.

Because prompting well isn’t just about writing a clever instruction, the success of its output depends on the context, judgment, and intent we provide within each prompt.

Today's modern marketers recognize that in order to leverage AI to its fullest potential, each prompt requires structure, nuance, and taste.

What This Guide Helps You Do

It’s here to serve you–a busy, multi-faceted, goal-driven, creative marketer. All day long, you are tasked with doing more; this guide helps you get to your final output in fewer attempts. We’ve spun our wheels on hundreds of variations of prompts and have curated this guide to help you get to your desired output with as few attempts as possible.

Write prompts that reflect real buyer behavior and GTM goals

Distill signals into messaging that makes sense to the human on the other end

Build reusable templates that scale without sacrificing taste

Collaborate with AI as a strategic partner, not a gimmick

Who This Is For

B2B marketing teams (ABM, Demand Gen, RevOps)

Early-stage founders building an account-based go-to-market motion

Curious revenue leaders, operators, and AI experimenters

Our Shared Philosophy

Young woman with long blonde hair wearing a black top and leather jacket, sitting indoors with green plants in the background.

Corrina Owens

"ABM isn't a campaign. It's a coordinated GTM motion. That motion works best when you act as the steward of account data and think in buying signals, not just personas or job titles."

Smiling woman with shoulder-length brown hair wearing a patterned navy blue top against a light background.

Elaine Zelby

"We've reached a breaking point with automation. AI allows us to scale 1:1 contextual messaging across channels. But only if we shift our mental model and use humans to guide the output with taste."

The Role of Taste in AI

We believe the best GTM teams will win not because they automate the most, but because they know what should be automated, and what still needs a human eye.

We call this the tastemaker layer.

AI can output content

But humans determine:

Is it emotionally resonant?

Is the framing correct for our buyer?

Does it respect timing, urgency, and nuance?

Are these the right channels to engage our audience?

Prompts reflect judgment.

And that judgment comes from you.

Choose Your Prompting Mode

The best prompts come when you know which job you want AI to do and you teach it how to think like you.

Before you start prompting,

ask yourself:

What type of person am I trying to emulate?

What am I trying to accomplish?

What is my desired output?

What information is required? What is not? Is that information inferred or verifiable?

Mode

Use AI to...

Prompt Example

Model

Thought Partner

Ideate, reframe, generate creative messaging, test POVs

How would you position this product for a skeptical CFO?

ChatGPT/Claude

Data Scientist

Analyze inputs, identify patterns, structure data output

Summarize the last 30 days of pricing page activity by key industry across our TAM.

ChatGPT

Content Marketer

Use to draft content outlines, research thought leadership topics

Using this template as tone and format guide, write a blog post on {{specific section}} of our Guide [Anchor Asset]

Claude

Social Media Manager

Analyze trends, draft or write posts in different tones, measure performance

Create 3 social posts for our upcoming product release: 1 in our corporate tone, one in our CEO tone, and one in our CMO tone

Claude/ChatGPT

Ad Specialist

Create ads across different channels and formats

Using this messaging doc, create the following: set of 12 headlines and 4 descrtiptons for Google responsive ads, 3 LinkedIn single image ads, 3 meta ads, and 3 reddit ads

ChatGPT/Gemini

PMM

Comb call recordings for insights, refine persona and industry specific messaging

Analyze all sales and cs calls over the past quarter and for the cybersecurity vertical tell me: what were the most common objections, questions, competitors mentioned, and areas of the product they were most excited about

Claude

Prompting Best Practices

Always write in first person, as if you're the individual making the ask

Use signal context to anchor every prompt

Tell the AI what format you want and what role it's playing

Prompt Framework

Element

Description

Example

Signal Context

What are we reacting to?

Competitor mention, re-engaged champion, website visit

AI Role

Who is the AI acting as?

AE, CMO, Revenue Analyst

Job

What are we asking it to do?

Write landing page copy, generate Slack message

Output Format

What structure should it use?

Bullet list, email sequence

Tone

What is the emotional nuance or audience mindset?

Direct, optimistic, skeptical, consultative

Prompt = Signal Context + AI Role + Job + Output Format + Tone

Prompt Library by GTM Use Case

Each prompt includes:

The signal you’re acting on

A real-world use case

A prompt that mirrors how a top-performing GTM team would frame it

1. Competitor Takeout Page

Signal Context: Competitor name mentioned on call

AI Role: AE who’s successfully handled this objection with competitor 100 times

Job: Generate side-by-side landing page and email sequence

Output Format:

Cheeky but respectful headline
3 buyer-relevant comparison points
Follow-up email sequence driving to landing page

Tone: Confident, consultative, never combative

Prompt:

You’re a rep who often hears “we’re evaluating [Competitor]” on sales calls. Write a side-by-side landing page that shows why customers switch to us. Be confident but not combative. Include a cheeky but respectful headline, and 3 comparison points that matter most to buyers like [Persona]. Generate email sequence copy that directs the user to the comparison page.

2. Win-Loss Pattern Analysis

Signal Context: Internal win-loss data over past 6 quarters

AI Role:Senior GTM data analyst

Job: Analyze patterns in wins and losses to build a predictive scoring model

Output Format:

Key win indicators
Loss patterns
Weighted scoring framework
Strategic summary for CMO/CRO

Tone: Analytical, strategic, consultative

Prompt:

You are a senior data analyst with deep experience in B2B SaaS and go-to-market strategy.Attached is a CSV with qualitative and quantitative win/loss data about our closed deals over the past 6 quarters. This includes industry, firmographic info, marketing source, deal length, buyer personas involved, engagement history, and whether the opportunity was closed won or lost. Your goal is to identify common patterns among won accounts so we can build a predictive model to prioritize future ICP accounts. This analysis should be actionable, strategic, and consultative, as if you’re presenting this to a CMO and CRO.

3. Multi-Signal Trigger and Generating Next Steps for Sales

Signal Context: ICP fit + past opp + former champion in seat

AI Role: ABM marketer reviewing signal data across target accounts

Job: Identify high-surging accounts and recommend tailored next steps

Output Format: Slack-ready message with:

One-sentence account insight
Suggested content
AE action recommendation

Tone: Clear, helpful, peer-to-peer

Prompt:

You are an account-based marketer analyzing signals across your ICP. You have engagement data from campaigns, website activity, and signals across your target accounts. Spot which accounts are showing high intent based on meaningful combinations of signal types. Then recommend one content suggestion and one outreach suggestion for each account, tailored to where they are in their buyer’s journey. Draft in a Slack ready message for the assigned account representative to take action on.

Real World Example: Slack Alerts with Taste

Corrina trained a GPT to summarize weekly ABM signals into Slack digests for reps. She fed it patterns from:

High-intent web visits
CRM opp stages + notes
Champion tracking

After 10+ iterations, it started to flag high-signal actions and suggest next steps like she would.

4. Industry and persona nuances for PMM messaging

Signal Context: Calls from past quarter from specific industry and/or persona

AI Role: PMM looking to refine messaging around a specific segment

Job: Identify nuanced language, needs, motivations, hesitations, and incentives of a specific audience

Output Format: Messaging doc with:

Section on identified themes and patterns for the specific segment
Bullets broken down by common: pain points, motivations, hesitations, and incentives
Specific value propositions to meet the needs of this audience

Tone: Comprehensive, authoritative, strategic

Prompt:

You are a product marketer analyzing customer and prospect calls with [segment]. Identify common themes related to this audience’s key pain points, motivations, hesitations, and incentives. Outline core themes and then present a list of bullets that highlight each of these specific dimensions. Then provide specific value propositions that should be used to address the needs of this segment. Draft a messaging doc for the PMM team to use in campaigns.

5. Nurture sequence for gated asset downloaders

Signal Context: Gated asset and leads who have filled out the form to download

AI Role: Lifecycle marketer

Job: Nurture and continue to education and engage hand raisers who have shown top-of-funnel intent but aren’t yet ready to talk to sales

Output Format: 5-email sequence with:

Engaging headlines that will get opens
Messaging that provides value and continues to build lead interest
CTAs that continue to push the lead through the funnel

Tone: Friendly, educational, inspiring

Prompt:

You are a lifecycle marketer who excels at nurturing top-of-funnel leads. Create a 5-email nurture series for leads that have downloaded [asset name + link to asset]. Each subject line should be intriguing, inspiring, and mysterious enough to get an open. The body copy should provide value specific to what this person would want to see next based on their industry, persona, and behavior, Make sure to include proof points, relevant statistics, and tactical practical information that is actionable. The CTAs should always be a soft push down funnel to another piece of content or event relevant to the individual.

In Conclusion

Prompting isn’t a passing skill. It’s becoming the operating system for how modern marketers think, create, analyze, and execute. The teams who win aren’t the ones who use the most AI, they’re the ones who know how to use it with intention.

This library gives you a starting point, but the real leverage comes when you adapt these prompts to your own style, your own buyers, and your own taste. The more you practice, the faster you’ll develop that instinctive “prompt muscle” that separates average operators from exceptional ones.

Use this guide to:

Shorten the distance between an idea and a polished execution

Turn raw signals into meaningful actions

Scale personalization without sacrificing nuance

Push your thinking further, faster, with AI as a true strategic partner

You now have the frameworks, modes, and structures to make every prompt sharper and every output more aligned with real GTM goals.

Now go build. Go experiment. And most of all: go prompt like a marketer.

Ready to Transform Your Prompting?

Start using these frameworks today in Tofu and see the difference in your AI-powered marketing workflows.

Get Started Learn More