How to use this
Run these five prompts in a single conversation thread, in order. Each prompt uses output from the previous step. Replace the [bracketed fields] with your specifics. Don’t skip steps — the chain compounds.
This grounds the entire chain. Everything downstream inherits what you define here. Spend the most time on this step — vague inputs produce vague campaigns.
Prompt
I'm planning a B2B marketing campaign. Here's the brief:
[Company name] sells [what you sell — be specific about the product/service] to [who buys it — job titles, company size, industry].
The goal of this campaign is [specific outcome: pipeline, signups, awareness, event registrations — pick one].
Timeline: [start date] to [end date]
Budget range: [if known, or "not yet defined"]
Channels and tactics that have worked before: [be specific — e.g., "LinkedIn ads targeting IT directors drove 40% of Q1 pipeline"]
Channels and tactics we've tried that didn't work: [be specific — e.g., "webinars consistently had <20 attendees and low conversion"]
Based on this brief, summarize back to me: who we're targeting, what we're trying to achieve, and what constraints we're working within. Flag anything that's too vague to act on.
Why this step matters
The last line — “flag anything that’s too vague” — is the most important part. It forces the LLM to push back on weak inputs before you build a campaign on top of them. Most teams skip straight to messaging and wonder why it feels generic.
↓
Before you write messaging, know what your audience is already hearing from everyone else. This step prevents you from saying the same things your competitors say.
Prompt
Here are competitors I know our buyer evaluates: [list 2-3 you're sure about]. Identify 2-3 more they're likely hearing from. For each one:
1. What's their primary marketing message right now?
2. What channels are they most visible on?
3. What claim or angle do they own in the buyer's mind?
Then tell me: what positioning gap exists that none of them are filling? What could we say that they can't or won't?
Before moving on
Check: does the competitive landscape match your experience? The LLM may miss niche competitors or overweight big names. Add any competitors it missed, then re-run. The positioning gap at the end is what you’ll build messaging around in step 4.
↓
Move from “who buys it” to “what are they dealing with right now.” This step builds the empathy layer that makes messaging land.
Prompt
Based on the brief and competitive scan, build a profile of our primary buyer. Cover:
1. Job title and reporting structure — who do they report to, who reports to them?
2. What is someone in this role typically measured on?
3. What's the biggest operational pain they're dealing with right now?
4. What have they already tried to solve it?
5. What would make them skeptical of a pitch from us?
6. What would make them forward our content to a colleague?
Write this as a narrative, not bullet points. I want to understand how this person thinks, not just what box they check on a form.
Where teams skip ahead
This is the step most teams skip or phone in. They jump from “we sell to VPs of Marketing” straight to messaging. The result is copy that describes the product instead of describing the buyer’s problem. Questions 5 and 6 are the ones that actually shape what you say.
↓
Now you have a brief, competitive context, and a buyer profile. This is the first step where you actually write marketing language — not before.
Prompt
Using the brief, competitive scan, and buyer profile from our conversation, build a messaging framework for this campaign:
1. Primary message — one sentence that captures what we want the buyer to believe after encountering this campaign. Not a tagline. A belief. (Example format: "Teams that [do X] outperform teams that [do Y] because [reason].")
2. Three supporting messages — each one addresses a different dimension of the buyer's problem (operational, strategic, emotional).
3. Proof points — for each supporting message, what evidence do we have? (Customer results, data, case studies, product capabilities.)
4. What we will NOT say — language, claims, or angles we should avoid based on the competitive scan and buyer skepticism.
Keep the language at a peer level. No superlatives, no "leading" or "cutting-edge." Write like you're explaining it to a smart colleague, not pitching a stranger.
Why "what we won't say" matters
The exclusion list is as important as the messaging itself. It prevents drift — the slow creep back toward generic language that happens when multiple people start creating assets from the framework. If everyone knows what’s off-limits, the campaign stays coherent across channels.
↓
Last step. You know the message — now decide where and how it shows up. The channel should follow the strategy, not the other way around.
Prompt
Based on everything in this conversation — brief, competitive landscape, buyer profile, and messaging framework — recommend a channel strategy for this campaign.
For each channel you recommend:
1. Why this channel reaches our specific buyer (not "because B2B companies use LinkedIn")
2. What format works best there (ad, organic post, email, landing page, event, etc.)
3. Which message from the framework fits this channel's context
4. One specific content idea with a working title
Also tell me: which channel should we start with if we can only do one? And which popular channel should we skip, and why?
Limit to 3-4 channels. More than that dilutes the campaign.
The constraint that matters
“Which channel should we skip” forces a real recommendation, not a safe list of everywhere you could be. The LLM will default to recommending every major channel unless you ask it to cut. The best campaigns say no to more channels than they say yes to.