Home / Category Playbooks
Category Playbooks

What to give it. What to ask for. How to fix it.

13 task categories. For each one: the right model, exactly what context you need to supply, what output to request, how to QA it, and what to do when the output is garbage.

Business Strategy

Browse prompts →
Best ModelChatGPT GPT-5.5 Thinking / Claude Opus 4.7
Default BrevityDetailed

What to Provide

  • Business model and current goal
  • Constraints: budget, timeline, resources
  • Market, competitors, target segment
  • Internal capabilities and team size

What to Ask For

  • Executive summary and diagnosis
  • Strategic options with risk trade-offs
  • Recommended path with rationale
  • Implementation plan + KPIs

Quality Checks

Output is specific to the provided business, not generic
Assumptions are clearly labeled where data was missing
No unsupported claims without source checks
Next actions are clear and executable this week
Common Failure

Generic output caused by weak context or vague success criteria.

Fix

Add audience, offer, examples, constraints, evidence, and the exact desired format before running.


Copywriting

Browse prompts →
Best ModelChatGPT GPT-5.5 / Claude Sonnet 4.6
Default BrevityStandard

What to Provide

  • The offer and exactly what it does
  • Target audience and their primary pain point
  • Proof points: reviews, stats, case studies
  • Tone, channel, length limits, and CTA
  • Top 2–3 objections you hear from prospects

What to Ask For

  • Copy variants organized by section (hook → body → proof → CTA)
  • Objection-handling language
  • Recommended test priority between variants

Quality Checks

Hook, offer, audience, proof, objection, and CTA are all addressed
Output is specific to the business — not template filler
Assumptions are clearly labeled
Next actions are clear
Common Failure

Generic output — sounds like every other DTC brand. Words like "crafted," "premium," "unleash," and "discover" everywhere.

Fix

Name a competitor URL and say "write copy that is deliberately different in tone." Add banned words and the exact voice reference.


Creative & Design

Browse prompts →
Best ModelCanva AI + ChatGPT GPT-5.5 / Claude Sonnet 4.6
Default BrevityStandard

What to Provide

  • Brand kit: colors (hex), fonts, logo usage rules
  • Asset type and exact dimensions
  • Visual do/don't references (URLs or descriptions)
  • Target audience and usage channel

What to Ask For

  • Creative brief with art direction
  • Layout and composition guidance
  • Asset specs and format notes
  • QA checklist before final export

Quality Checks

Output is specific to the provided brand and context
Assumptions are clearly labeled
No unsupported claims
Next actions are clear and executable
Common Failure

Output is generic creative direction with no brand specificity. Could be for any business.

Fix

Add your exact hex codes, a reference image URL, the channel, and one "this is what we never do" rule. Specificity breaks genericness.


Data & Analytics

Browse prompts →
Best ModelChatGPT GPT-5.5 Thinking / Gemini 3.1 Pro Preview
Default BrevityDetailed

What to Provide

  • Dataset or schema description
  • KPI definitions and target benchmarks
  • Date range and segment of interest
  • The business decision this analysis should inform

What to Ask For

  • KPI table with period comparisons
  • Findings and interpretation
  • Recommended action based on data
  • Caveats and data quality notes

Quality Checks

Analysis answers the specific business question, not a generic summary
Assumptions about missing data are explicitly flagged
Claims are tied to actual numbers in the dataset
Next action is concrete and tied to a specific metric
Common Failure

Output describes what the data shows but not what to do. All observation, no decision.

Fix

End your prompt with: "What is the single most important action to take, and by when?" Forces a decision, not a description.


Best ModelChatGPT GPT-5.5 / Claude Sonnet 4.6
Default BrevityStandard

What to Provide

  • Goal in one sentence
  • Context: who, what, where, why
  • Constraints and deadline
  • One example of the output you want

What to Ask For

  • Structured answer with clear assumptions labeled
  • Recommendations with short rationale
  • Examples where relevant
  • Specific next steps

Quality Checks

Output addresses the actual goal, not a paraphrase of the goal
Assumptions are labeled
No unsupported claims
Next step is clear
Common Failure

Output is vague because the goal was vague. Garbage in, garbage out applies more here than anywhere.

Fix

Start over with: "My goal is [X]. I want the output to be [format]. Here's an example: [Y]." Three sentences. Runs every time.


Best ModelGemini 3.1 Flash Image Preview / Canva AI / ChatGPT GPT-5.5 for prompt refinement
Default BrevityExact Spec

What to Provide

  • Subject with specific detail (not "a woman" — "a 30s chef in a white apron")
  • Style reference: photographer, film, art movement, or direct URL
  • Composition, aspect ratio, and brand colors
  • Lighting type and negative prompts (what to exclude)

What to Ask For

  • Production-ready prompt with all six elements: subject, style, composition, lighting, camera, environment
  • Aspect ratio specified
  • Negative prompt list

Quality Checks

Prompt includes camera/composition, motion, lighting, and aspect ratio
Negative prompts are present
Subject description is specific enough to produce consistent results
Style reference is grounded (not just "cinematic")
Common Failure

Output looks technically fine but could belong to any brand. No visual identity carries through.

Fix

Add your exact brand hex colors to the prompt. Add "consistent with [reference URL]." Add one negative prompt: "watermark, text overlay." Three changes, dramatically more on-brand.


Marketing

Browse prompts →
Best ModelChatGPT GPT-5.5 / Claude Sonnet 4.6
Default BrevityStandard

What to Provide

  • Business, offer, and target audience
  • Budget and target geography
  • Channel (Meta, Google, email, organic) and current results
  • Competitor examples that are working
  • Primary success metric (ROAS, CPA, leads, revenue)

What to Ask For

  • Campaign strategy with creative angles
  • Channel setup and budget allocation
  • Audience definition and targeting logic
  • KPIs with expected ranges
  • Next 3 actions to launch

Quality Checks

Hook, offer, audience, proof, objection, and CTA are addressed
Strategy is specific to the business and channel, not generic
Assumptions are labeled
Next actions are executable this week
Common Failure

Campaign plan reads like a textbook. "Build awareness, nurture leads, convert." Useless for execution.

Fix

Add: "My current cost-per-lead is $X. My goal is $Y. My top-performing ad shows [description]. Beat it." Now the model has a target and a reference. Output gets specific.


Productivity

Browse prompts →
Best ModelChatGPT GPT-5.5 / Claude Sonnet 4.6
Default BrevityStandard

What to Provide

  • Current workflow (step by step, as it actually runs)
  • Tools involved and people responsible
  • The bottleneck: where it breaks or slows
  • Frequency (daily, weekly, per-client) and approval rules

What to Ask For

  • Workflow map with owners
  • SOP or checklist
  • Automation opportunities with tool suggestions
  • RACI if multiple people are involved
  • Maintenance cadence

Quality Checks

Output matches the actual tools your team uses, not generic alternatives
Assumptions are labeled
Each step has a clear owner
Next actions are executable without additional design decisions
Common Failure

SOP is written for a generic business with generic tools. Doesn't match how your team actually works.

Fix

Name the actual tools: "We use Notion, Slack, and Stripe. Not Asana, not HubSpot." Add "Write the SOP as if you're handing it to a new hire who has never done this job." Forces operational specificity.


Prompt Engineering

Browse prompts →
Best ModelChatGPT GPT-5.5 Thinking / Claude Opus 4.7
Default BrevityStandard

What to Provide

  • Target model and tool (Claude, GPT-5.5, Gemini — each needs different framing)
  • Objective: what the prompt should produce every time
  • Inputs that will vary across runs
  • Constraints: length, format, banned words
  • One example of ideal output

What to Ask For

  • Reusable prompt template with variables marked [like this]
  • Instructions written inside the prompt (not just around it)
  • Output format specification
  • Validation tests: 3 different inputs to try

Quality Checks

Prompt produces consistent output across different inputs
Variables are clearly marked and documented
Output format is specified inside the prompt, not just assumed
Tested against at least 2 different inputs before trusting it
Common Failure

Prompt works once, fails on the second input. Built around the example, not the objective.

Fix

Test with the edge case first: the hardest, most unusual input you'll ever feed it. If it breaks, fix the prompt for that case. Don't fix for the easy case and assume it generalizes.


SEO & Content

Browse prompts →
Best ModelChatGPT GPT-5.5 + Perplexity Sonar for current research
Default BrevityDetailed

What to Provide

  • URL of the page being optimized or created
  • Target keyword and search intent (informational, commercial, transactional)
  • Target audience and location
  • Competitor URLs ranking for the same keyword
  • Internal links to include and products/services to reference

What to Ask For

  • SEO brief or full content draft
  • Outline with H2/H3 structure
  • On-page elements: title tag, meta description, H1
  • FAQ section and schema markup suggestions
  • Internal linking opportunities

Quality Checks

Search intent, internal links, FAQ, and on-page elements are included
Article sounds like a real expert, not SEO filler
Assumptions are labeled
CTA is present and natural
Common Failure

Article is technically on-topic but says nothing new. No original angle. Will not rank above what already exists.

Fix

Add your stance: "Our take on this topic is [X] — which disagrees with the common advice to [Y] because [Z]." Give the model an argument to make, not just a topic to cover.


Sales & E-commerce

Browse prompts →
Best ModelChatGPT GPT-5.5 Thinking / Claude Sonnet 4.6
Default BrevityDetailed

What to Provide

  • Website or store URL
  • Product or service and its core promise
  • Target audience and funnel stage
  • Analytics: traffic, conversion rate, cart abandonment
  • Current conversion blocker (what you believe is wrong)

What to Ask For

  • Conversion diagnosis with prioritized fixes
  • Copy and UX recommendations per page
  • A/B test plan with specific variants to run
  • KPI impact estimate for top fixes

Quality Checks

Fixes are specific to the actual page, not generic CRO advice
Priority order is based on impact, not effort
Assumptions are labeled
Next actions are executable without a developer for the top 3
Common Failure

Diagnosis is a list of everything that could be improved. No priority order. Useless for a founder who has 4 hours.

Fix

End the prompt with: "If I can only fix one thing before next week's campaign launch, what is it and what exactly should it say?" Forces a decision, not a wishlist.


Social Media

Browse prompts →
Best ModelChatGPT GPT-5.5 / Claude Sonnet 4.6
Default BrevityStandard

What to Provide

  • Brand voice reference (a post or two that worked)
  • Target audience and platform (LinkedIn ≠ Instagram ≠ TikTok)
  • Goal: followers, website clicks, DMs, sales
  • Offer or topic being promoted
  • Posting cadence and visual reference

What to Ask For

  • Platform-ready hooks and captions
  • Posting sequence with dates/timing
  • Creative direction for visuals
  • CTA variants for each post

Quality Checks

Hook, offer, audience, proof, objection, and CTA are all addressed
Posts match the platform's native content style, not a generic template
Voice matches the provided brand reference
Next action is clear (schedule, design, post)
Common Failure

Posts sound like a corporate newsletter. No personality. No hook that stops the scroll.

Fix

Paste in your best-performing post and say "match this voice exactly." Then add: "The first word cannot be 'I,' 'We,' or 'Our.'" Watch the hooks improve immediately.


Best ModelHiggsfield / Gemini / ChatGPT GPT-5.5 for prompt refinement
Default BrevityExact Spec

What to Provide

  • Scene concept and hook (first 2 seconds matter most)
  • Platform and duration (Reels = 15s, YouTube Short = 60s)
  • Subject, camera movement, and intended emotion
  • Script or voiceover if applicable
  • Negative prompts: what the video should never do or show

What to Ask For

  • Production-ready video prompt with scene, motion, shot sequence
  • Audio/VO notes and pacing
  • Platform framing (vertical vs horizontal vs square)
  • Negative prompt list

Quality Checks

Prompt includes camera movement, motion type, lighting, and aspect ratio
Negative prompts are present
Hook is described at the scene level, not just "engaging opening"
Platform and duration are specified
Common Failure

Generated video looks nothing like the concept. Too abstract. No direction on motion or camera.

Fix

Describe the camera move first: "Slow zoom in on [subject] from wide to medium. No cuts." Then describe the subject. Camera direction before subject description consistently produces tighter results.


Ready to run these workflows?

Open the Vault — 1,096 prompts free