Why this is hard to get right
The Brief That Kept Bouncing Back
Maya is a performance marketing manager at a mid-sized SaaS company. Her team runs retargeting campaigns across Google Display and LinkedIn, and every quarter she faces the same problem: the creative briefs she hands to designers and copywriters come back with follow-up questions.
"What size should the headline be?" "Can we mention the competitor?" "What's the actual offer here?" Three rounds of back-and-forth later, the campaign launches two weeks late.
Maya knows the campaign should work. The audience is warm — these are pricing-page visitors who almost converted. The product has real proof points. The budget is there. But the brief keeps failing before the ads ever run.
She tried asking AI tools for help, but the outputs were frustratingly generic. She'd type something like "write retargeting display ads for our SaaS tool" and get back aspirational headlines like "Grow Your Business Today" — copy that could belong to any company in any industry.
The problem wasn't the AI. The problem was the prompt.
What Maya was missing was structure and specificity. She needed to tell the AI the exact audience segment — not just "warm traffic" but "visitors who viewed the pricing page and didn't start a trial." She needed to specify that this audience had seen the product, understood the value proposition broadly, and was hesitating — probably on price, trust, or timing.
She also needed to define hard constraints upfront: character limits for 300x250 and 728x90 placements, a ban on competitor mentions, no discount offers, and a clear 30-day conversion goal.
Once Maya restructured her prompt with those details, the output changed completely. The AI returned headlines that addressed pricing hesitation directly. The proof points were specific and credible. The visual concepts mapped to real placement sizes. She handed the brief to her designer with zero follow-up questions.
The campaign launched on time. The trial-start rate climbed 22% in the first three weeks.
The lesson Maya learned applies broadly: retargeting briefs fail when they treat all visitors the same. Warm traffic is not cold traffic. Pricing-page visitors are not blog readers. A good brief captures that intent signal and uses it to drive every creative decision — the offer angle, the proof points, the visual hierarchy, and the tone.
AI can produce genuinely useful retargeting creative briefs, but only when you give it the behavioral context, placement specs, and guardrails it needs. That specificity is what separates a brief that executes cleanly from one that bounces between teams for two weeks.
Common mistakes to avoid
Treating All Retargeting Audiences as One Segment
Pricing-page visitors and blog readers require completely different messaging. Lumping them together produces copy that addresses no one's specific hesitation. Always tell the AI exactly which behavioral trigger defines your audience — page visited, time on site, cart abandon — so the copy can address real objections rather than generic benefits.
Skipping Character Limits and Placement Specs
Display ads have hard character caps that vary by size. If you don't specify placement dimensions and limits upfront, the AI writes copy at the wrong length. You'll spend time cutting headlines and rewriting descriptions. Always include character limits and placement sizes — 300x250, 728x90, 160x600 — in the prompt itself.
Leaving the Offer Open-Ended
Asking for 'a compelling offer' without constraints forces the AI to guess. It often defaults to discounts, free trials, or aggressive CTAs that your legal or marketing team will reject. Specify what you can and cannot offer — for example, 'no discounts, no competitor mentions, no urgency claims' — and name the specific CTA you're allowed to use.
Omitting Proof Points From the Prompt
Without real proof points, the AI invents vague credibility claims like 'trusted by thousands of users.' These claims are weak and often inaccurate. Supply 2–3 real data points — a specific time-savings percentage, a customer count, a G2 rating — so the copy stays credible and defensible in any compliance review.
Setting No Conversion Goal or Success Metric
A prompt without a stated goal produces copy that sounds good but optimizes for nothing. Tell the AI the specific metric you're targeting — trial starts, demo bookings, return visits — and the timeframe. This shapes word choice, CTA strength, and offer framing in ways that a vague 'improve conversions' instruction never will.
Writing the Brief for the Wrong Reading Level
Retargeting ads compete for attention in milliseconds. If your prompt doesn't specify tone — 'clear and practical, not clever or hype-driven' — the AI often writes witty or abstract copy that performs poorly on banner placements. Specify tone explicitly and ban language patterns (hype words, jargon, rhetorical questions) that slow comprehension.
The transformation
Write some retargeting ads for my company and make them convert better.
You’re a performance copywriter creating a retargeting **display ad creative brief**. **Context:** We sell a $49/month B2B SaaS time-tracking tool. Visitors viewed the pricing page but didn’t start a trial. 1. Write **6 headlines (≤30 chars)** and **6 descriptions (≤90 chars)**. 2. Provide **2 offer angles** (no discount) and **3 proof points**. 3. Recommend **2 visual concepts** for 300x250 and 728x90. 4. Keep tone **clear, confident, and practical**. **Goal:** Increase trial starts by **20%** in 30 days. Avoid competitor mentions and hype.
Why this works
Behavioral Context Shapes Copy
The After Prompt specifies 'visitors viewed the pricing page but didn't start a trial.' That single behavioral signal tells the AI this audience understands the product and is hesitating — not unaware. Every headline, proof point, and offer angle then addresses hesitation, not awareness, which is the correct job for retargeting copy.
Numbered Tasks Prevent Omissions
The After Prompt breaks the deliverable into four numbered steps: headlines, descriptions, offer angles, and visual concepts. This structure forces the AI to complete each output type in a single pass. Without it, AI tools often skip proof points or visuals entirely, requiring follow-up prompts to fill the gaps.
Character Limits Eliminate Rework
The After Prompt sets '6 headlines (30 chars)' and '6 descriptions (90 chars)' — limits that match real Google Display Network specs. This means the output is placement-ready rather than needing manual trimming. Designers and copywriters receive copy they can use immediately, reducing revision cycles.
Guardrails Define What's Off-Limits
The After Prompt explicitly bans competitor mentions and hype language. These negative constraints are as important as positive instructions. Without them, the AI defaults to patterns that are legally risky or brand-inconsistent. Explicit exclusions protect the brief from unusable copy before it's ever written.
A Quantified Goal Anchors the Output
The After Prompt states 'Increase trial starts by 20% in 30 days.' That concrete goal signals to the AI that this brief needs direct-response copy, not brand copy. It shifts the output toward specificity, urgency calibration, and action-oriented language — the hallmarks of high-converting retargeting creative.
The framework behind the prompt
The Strategy Behind Retargeting Creative Briefs
Retargeting advertising operates on a fundamentally different psychological model than prospecting. Where cold-traffic ads build awareness and desire, retargeting ads must resolve hesitation and doubt among audiences who already know what you offer.
The dominant framework in direct-response advertising — AIDA (Attention, Interest, Desire, Action) — applies differently at this stage. Retargeting audiences have already passed through Attention and Interest. Your creative brief must focus almost entirely on Desire reinforcement and Action facilitation.
Research from the Fogg Behavior Model helps explain why retargeting fails when motivation alone is insufficient. Conversion requires three elements: motivation (the buyer wants the outcome), ability (the action is easy), and a prompt (a trigger arrives at the right moment). Retargeting ads serve as that prompt — but only when the creative removes friction rather than adding noise.
This is why audience segmentation by behavioral trigger is not optional in retargeting briefs. A visitor who spent 8 minutes on your pricing page and returned twice is in a different cognitive state than someone who bounced after 30 seconds. The copy, offer, and proof points must map to the specific hesitation pattern of each segment.
The creative brief as a strategic document — not just a production handoff — is a principle from agency practice. A strong brief encodes the audience's intent signal, the specific objection the creative must address, the proof point that resolves that objection, and the precise action the ad requests. When all four elements align, conversion rates improve not because the creative is clever but because it is situationally precise.
Character limits in display advertising also carry strategic weight. The cognitive load model suggests that attention-constrained formats (banner ads competing with page content) reward radical clarity over persuasive complexity. Shorter, simpler copy consistently outperforms longer copy in standard display placements — which is why this prompt structure enforces hard character caps at the brief stage, not the revision stage.
Prompt variations
You're a direct-response copywriter creating a retargeting display ad brief for an e-commerce brand.
Context: We sell premium leather bags at $180–$340. Visitors added a product to cart but did not complete checkout. Average session lasted 4 minutes.
- Write 6 headlines (30 chars max) and 6 descriptions (90 chars max) for Google Display.
- Provide 2 urgency angles that do not use countdown timers or false scarcity.
- Recommend 2 visual concepts for 300x250 and 300x600 placements.
- Suggest 3 proof points drawn from quality, craftsmanship, or customer satisfaction.
Goal: Recover 15% of abandoned carts within 7 days of session.
Constraints: No discount codes. No 'limited time' language. Tone is calm, confident, and premium — not pushy.
You're a B2B performance marketer writing a LinkedIn retargeting ad brief.
Context: We offer project management software at $29/user/month. The audience is users whose free 14-day trial expired without upgrading. Company size: 50–500 employees. Roles: operations managers and team leads.
- Write 4 LinkedIn single-image ad headlines (70 chars max) and 4 body copy variations (150 chars max).
- Suggest 2 re-engagement angles focused on value missed, not features lost.
- Recommend 1 visual concept appropriate for a professional LinkedIn feed.
- Provide 3 proof points using customer outcomes, not product specs.
Goal: Reactivate 10% of expired trial users into paid seats within 30 days.
Constraints: No discount offers. No 'your trial has expired' language — frame it as a return, not a reminder of loss. Tone: direct, peer-to-peer, not salesy.
You're a senior performance copywriter at a digital agency creating a reusable retargeting display ad brief template.
Context: This template serves multiple B2B SaaS clients. Each client targets pricing-page visitors or free-trial drop-offs. Clients range from $29/month productivity tools to $500/month enterprise analytics platforms.
Build a brief template that includes:
- A 5-question intake checklist the account manager completes before briefing the AI.
- Placeholder copy structures for 6 headlines (30 chars) and 6 descriptions (90 chars) with annotation on where client proof points slot in.
- 3 universal offer angle frameworks (no discount, no urgency) that work across B2B SaaS price points.
- A visual concept recommendation matrix for 300x250, 728x90, and 160x600 formats.
Goal: Reduce brief-to-execution time from 5 days to 2 days across all client accounts.
Constraints: Template must work without client-specific data until the intake checklist is complete. Tone guidance must be adjustable per brand voice.
You're a retention copywriter creating a retargeting display ad brief for a customer success team.
Context: We run a $79/month HR onboarding platform. The audience is paying customers who have not logged in for 21 or more days and show low feature adoption. These users are at churn risk but are still subscribers.
- Write 6 headlines (30 chars max) and 6 descriptions (90 chars max) focused on re-engagement, not upsell.
- Provide 2 reactivation angles that highlight value the user is currently missing — not features they haven't tried.
- Recommend 2 visual concepts for 300x250 that feel helpful, not alarming.
- Suggest 3 proof points from customer success metrics, peer benchmarks, or time-savings data.
Goal: Increase 30-day login rate by 25% among at-risk accounts.
Constraints: No 'you haven't logged in' language. No upgrade offers. No urgency framing. Tone: warm, practical, and supportive.
When to use this prompt
Marketing Managers running retargeting tests
Create a complete brief for your next 2–4 week retargeting sprint, including angles and proof points.
Product marketers launching a new feature upgrade
Retarget pricing-page visitors with feature-specific benefits and clear visuals that match intent.
Sales teams supporting pipeline acceleration
Align ads to common late-stage objections and drive more demo or trial starts from warm traffic.
Customer success teams reducing churn risk
Retarget downgraded or inactive users with reactivation messaging that avoids discounts and stays helpful.
Pro tips
- 1
Specify the retargeting segment so the AI matches intent and objections.
- 2
Add 2–3 proof assets so your copy stays credible, not generic.
- 3
Define placements and character limits so you don’t rewrite ads for every size.
- 4
Clarify what you can’t offer so the model avoids unusable promos and claims.
Most retargeting prompts define one behavioral trigger. Advanced briefs layer two or three signals to narrow the audience and sharpen the message.
For example, instead of 'visited the pricing page,' try: 'visited the pricing page, viewed the enterprise tier, and returned within 48 hours.' That layered signal suggests a high-intent buyer who is comparison shopping — not casually browsing.
When you include that level of specificity in your prompt, the AI can address comparison-stage objections directly: trust signals, contract flexibility, onboarding support, and peer-company references. The copy moves from 'here's why we're good' to 'here's why we're the right choice over what you're already considering.'
Three layering techniques to try:
- Combine page visited with session duration (3+ minutes signals genuine evaluation)
- Add recency window (visited in last 3 days vs. last 30 days changes urgency calibration)
- Include return visit data (second visit to pricing page signals active comparison)
Each additional signal narrows the audience and expands the precision of your copy brief. The AI doesn't need to guess at intent — you've told it exactly where the buyer is in the decision process.
A single retargeting brief often needs to serve four or five placement sizes simultaneously. Each size has different character limits, visual real estate, and reading context.
Common display placements and their copy constraints:
- 300x250 (Medium Rectangle): Most common placement. Headline up to 30 chars, description up to 90 chars. Competing for attention in sidebars — copy must be direct and scannable.
- 728x90 (Leaderboard): Horizontal banner, typically at page top. Very limited vertical space. Headline dominates — description often truncated. Prioritize the headline.
- 160x600 (Wide Skyscraper): Tall, narrow format. Works well for sequential visual storytelling. Two-line copy with a logo performs better than a dense text block.
- 300x600 (Half Page): Premium placement with room for longer copy and stronger visual. You can use a short testimonial or a 2-line proof point here.
When briefing across all sizes in one prompt, specify each placement by name and list its character limits separately. Ask the AI to produce a version for each format rather than asking it to 'adapt the copy' after the fact. Front-loading this requirement saves one full revision cycle.
Many teams assume retargeting requires a discount to work. That's false — and in many categories (premium products, enterprise software, regulated industries), discount offers actively damage brand perception.
Here are three offer angle frameworks that convert without discounting:
1. The Risk Removal Angle Focus on what the buyer doesn't lose by trying. Extended trial periods, money-back guarantees, no-contract commitments, and free onboarding support all reduce perceived risk without cutting price. Copy example: 'Full access. No card required. Cancel any time.'
2. The Peer Proof Angle Use social proof from companies or roles similar to the viewer's context. 'Teams like yours at Acme Co. onboarded in 2 days' is more persuasive than a generic testimonial. Name the industry, company size, or job role whenever possible.
3. The Asymmetric Value Angle Show the cost of not acting in concrete terms. If your tool saves 5 hours per week, calculate the annual cost of that lost time at a reasonable hourly rate. The 'cost of inaction' frame reframes price as an investment comparison — not a spend.
When you specify which angle to use in your prompt, the AI produces tighter, more coherent copy than when you leave the angle open.
When not to use this prompt
When This Prompt Is Not the Right Tool
This prompt structure is built for warm, intent-rich retargeting audiences. It breaks down in several scenarios:
- Cold prospecting campaigns: If your audience has never visited your site or engaged with your brand, this brief's assumption of existing awareness will produce copy that references context the audience doesn't have. Use a top-of-funnel creative brief instead.
- Brand awareness display campaigns: If your goal is reach and recall rather than direct conversion, the tight CTA structure and proof-point framing in this brief will push you toward performance copy when you need brand-building copy. Those are different disciplines.
- Highly regulated industries: In pharma, finance, or legal services, the AI-generated proof points and offer angles in this brief require mandatory legal review before any media placement. Don't treat AI output as compliance-ready copy in regulated categories.
- Very early-stage products without data: If you have no real proof points — no customer count, no ratings, no outcome metrics — this brief will either produce fabricated specifics or fall back to vague claims. Build your proof point library before running this prompt.
In these cases, adapt the structure to the actual campaign goal rather than forcing retargeting framing onto a different creative objective.
Troubleshooting
Headlines are consistently too long despite specifying character limits
Move the character limit directly into the task line, not just the context block. Write: 'Write 6 headlines. Each must be 30 characters or fewer — count every character including spaces.' Then ask the AI to output the character count in parentheses after each headline so you can verify compliance without manual counting.
The AI writes generic proof points like 'trusted by thousands of businesses'
The prompt is missing real data. Add at least two specific numbers to your context block before asking for proof points — for example, 'Our customers report 40% faster onboarding and a 4.7 G2 rating across 800 reviews.' The AI cannot invent credible specifics; you must supply the raw data and ask it to shape them into usable copy.
Offer angles keep defaulting to discounts even when the prompt says no discounts
Repeat the constraint inside the numbered task, not just in the guardrails block. Write: 'Provide 2 offer angles. Do not suggest discounts, free extensions, or price reductions of any kind.' Also add: 'If you cannot suggest an angle without a discount, explain why and propose an alternative.' This prevents silent non-compliance.
Visual concept recommendations are too abstract to execute
Ask for visual concepts in a concrete production format: 'Describe each visual concept in 2 sentences. Include the foreground subject, background treatment, and the single emotion it should evoke.' Abstract outputs like 'a clean, modern feel' are useless to designers. Structured descriptions give the creative team something they can actually execute.
The output mixes B2C and B2B tone, producing copy that feels inconsistent
Add a two-sentence voice rule to your prompt's tone block. For example: 'Write for operations managers at 100-person companies. Use peer-to-peer language — direct and collegial, not promotional.' Without a named audience role and relationship frame, the AI averages across tones and produces copy that fits neither context.
How to measure success
How to Evaluate the Output Quality
A strong retargeting creative brief from this prompt should pass a practical checklist before you hand it to your creative team.
Structural completeness:
- All requested deliverables are present — headlines, descriptions, offer angles, proof points, and visual concepts
- Character counts fall within the specified limits for each placement
- Visual concepts name a specific foreground element, not just a mood or aesthetic
Copy quality signals:
- Headlines address a specific hesitation, not a generic benefit
- At least one proof point uses a real number (percentage, count, rating)
- Offer angles do not default to discounts if discounts were excluded
- Tone is consistent across all six headline variations — no tonal drift from direct to hype
Usability for execution:
- A designer could start work on the visual concept without a follow-up question
- A copywriter could populate an ad platform with the copy in under 15 minutes
- Legal or compliance could identify any claims that need sourcing
Red flags to reject and re-prompt:
- Any headline that could belong to a competitor without change
- Proof points that use "many," "most," or "thousands" without a specific number
- Visual concepts described only as "clean and modern"
Now try it on something of your own
Reading about the framework is one thing. Watching it sharpen your own prompt is another — takes 90 seconds, no signup.
Build a complete retargeting brief — audience segment, offer angles, proof points, and placement-ready copy — in one pass.
Try one of these
Frequently asked questions
As specific as your data allows. At minimum, name the behavioral trigger — page visited, trial expired, cart abandoned. Better prompts also include time window (visited in last 7 days), session depth (viewed pricing page), and any known job role or company size. The more intent signals you include, the more precisely the AI can match objections to copy.
Yes, but adjust the character limits and format requirements to match each platform. LinkedIn single-image ads use 70-char headlines and 150-char body copy. Meta feed ads use 40-char headlines and 125-char primary text. Swap the placement specs in the prompt, keep the audience context and constraints, and the structure works across channels.
Add a constraints block at the end of your prompt that lists every restriction explicitly. For example: 'No discount language. No competitor mentions. No claims that cannot be sourced. No urgency framing.' AI tools follow negative instructions well when they are specific. Vague guardrails like 'stay on brand' produce inconsistent results.
Use any verifiable data you have internally: aggregate product metrics (users save X hours per week), satisfaction scores from NPS surveys, review platform ratings (G2, Capterra), or feature adoption rates. You don't need a formal case study. A single specific number — '93% of users report faster onboarding' — outperforms a vague claim every time.
Restate the character limit at the top of the numbered task, not just in the context block. For example: 'Write 6 headlines. Each headline must be 30 characters or fewer — count carefully.' Repeating the constraint where the task appears reduces length errors significantly. You can also ask the AI to output the character count next to each headline.
Yes, always. Append a short voice note after your tone instruction. For example: 'We write in second person, active voice, with no jargon. We never use exclamation points.' Two to four specific voice rules outperform a general label like 'professional tone' because they give the AI patterns to follow, not a category to guess at.
Six headlines and six descriptions is a practical ceiling for a single prompt. Asking for more than eight of any element tends to produce padding and repetition. If you need additional variations, run a follow-up prompt asking the AI to rework the weakest three with a different angle — this produces sharper alternatives than asking for twelve upfront.
Yes, with two adjustments. Native placements use longer headlines (up to 100 characters) and lean on curiosity-driven framing rather than direct response. Swap the character limits, replace 'clear and practical' with 'intriguing but credible,' and specify that the ad should feel like editorial content — not a promotional banner. Everything else in the structure holds.