25 ChatGPT-5.5 Prompts for Small Business Owners: Marketing, Operations, Finance, and Customer Service Automation

25 ChatGPT-5.5 Prompts for Small Business Owners: Marketing, Operations, Finance, and Customer Service Automation

Updated July 2026 — This long-form guide delivers 25 battle-tested, copy-paste-ready ChatGPT-5.5 prompts designed for small business owners, founders, and operations managers who want practical automation across Marketing & Content, Operations & Productivity, Finance & Accounting, and Customer Service. Each prompt includes context variables, recommended model settings, example outputs, and step-by-step integration advice so you can deploy them immediately with ChatGPT-5.5 or in your platform workflow (Zapier, Make, Google Apps Script, or an in-house automation layer).

These prompts are written for advanced ChatGPT power users and professionals who need reliable, reproducible results. Use them as templates, refine tone and outputs for your brand, and combine with tools and API function calls where appropriate. Keywords covered: ChatGPT prompts small business, AI prompts entrepreneurs, business automation prompts, ChatGPT for SMBs.

Table of Contents

• Overview: How to use these prompts with ChatGPT-5.5
• Section 1 — Marketing & Content (9 prompts)
• Section 2 — Operations & Productivity (6 prompts)
• Section 3 — Finance & Accounting (4 prompts)
• Section 4 — Customer Service (6 prompts)
• Prompt Engineering Best Practices & Model Settings
• Summary Table: Quick reference
• Integration Tips: APIs, Zapier, Google Sheets, and Function Calls
• Further Reading & Internal Resources

Overview: How to use these prompts with ChatGPT-5.5

ChatGPT-5.5 in July 2026 typically offers: a larger context window (200k+ tokens in many deployments), multimodal inputs (text + image + document), and robust tool/function calling integrations for external systems (scheduling, CRMs, spreadsheets). The prompts below assume you can:

• Send a system message (assistant persona and constraints)
• Adjust temperature, top_p, and max_tokens
• Receive structured outputs (JSON, CSV, Markdown tables) for automated ingestion
• Use chain-of-thought style breakdowns for SOPs and step-by-step tasks

Before using any prompt, set a system prompt describing long-term constraints (brand voice, prohibited content, data retention rules). Example system message: “You are the virtual growth operator for {{BUSINESS_NAME}}. Write concise, fact-driven, and action-oriented outputs. When asked for structured data, return strictly valid JSON with no extraneous commentary.” Refer to platform docs for how to pass system messages to ChatGPT-5.5.

This collection is optimized for small business workflows and integrates with common SMB stacks. If you’re implementing end-to-end automation, consider pairing these prompts with webhook-based triggers and simple validation scripts. For tactical examples and platform-specific connectors, see ChatGPT prompts small business and business automation prompts.

Section 1 — Marketing & Content (Social Media, Email Campaigns, SEO)

Goal: Create repeatable, brand-aligned content and campaign assets that save time and scale. Use these prompts for weekly content calendars, SEO optimizations, ad copy, and email campaigns. Recommended model settings for this section: temperature 0.2–0.5 for consistent copy, max_tokens 600–1200 depending on length required.

Prompt 1 — 30-Day Social Media Calendar (Multi-Platform)

Purpose: Generate a platform-specific 30-day content calendar (LinkedIn, Instagram, X, Facebook, TikTok) with post copy, hashtags, suggested assets, and posting times aligned to a small business’s timezone and audience.

Context variables (replace before running):

• {{BUSINESS_NAME}} — brand name
• {{INDUSTRY}} — e.g., specialty coffee, B2B SaaS, boutique retail
• {{AUDIENCE}} — primary audience segments (age, role, interests)
• {{BRAND_VOICE}} — e.g., professional, playful, direct
• {{TIMEZONE}} — e.g., America/Los_Angeles
• {{PRIMARY_GOAL}} — brand awareness, lead generation, sales

System: You are a social media content strategist for {{BUSINESS_NAME}}. Output JSON only.

User:
Create a 30-day social media calendar for {{BUSINESS_NAME}} (industry: {{INDUSTRY}}). For each day, produce one post per platform: LinkedIn, Instagram (feed), X (Twitter), Facebook, and 1 short-form idea for TikTok. For each entry include:
- date (YYYY-MM-DD)
- platform
- post_copy (<= 280 chars for X, <= 2200 for Instagram)
- suggested_image_or_video (brief description)
- hashtags (3-8)
- CTA (one line)
- best_post_time (local to {{TIMEZONE}})
- performance_estimate (expected engagement: Low/Medium/High, with reason)

Constraints:
- Tone: {{BRAND_VOICE}}
- Target audience: {{AUDIENCE}}
- Primary goal: {{PRIMARY_GOAL}}
- Avoid making unverifiable claims. Use action verbs and a single CTA per post.

Return strictly valid JSON array of 30*5 objects.

Why it works: Structured JSON allows direct upload to scheduling tools (Buffer, Later) or into a Google Sheet. Example production note: schedule posts within platform-recommended peak times for small businesses (LinkedIn: Tue–Thu 9–11am; Instagram: Wed 11am & Fri 10am). Use A/B tests on captions days 7–10 and 21–24.

Prompt 2 — High-Converting Email Campaign Sequence (5 emails)

Purpose: Create a 5-step email nurture sequence (welcome, value, social proof, offer, re-engage) with subject lines optimized for open rates.

System: You are an email marketing strategist for {{BUSINESS_NAME}}. Return a JSON object with 5 emails.

User:
Write a 5-email nurture sequence for new subscribers to {{BUSINESS_NAME}}. Include:
- email_number (1-5)
- subject_line (A/B variant)
- preheader (<= 90 chars)
- body_html (clean, mobile-friendly HTML)
- plain_text_version
- suggested_send_delay (e.g., immediately, +2 days)
- primary_metric (open/CTR/conversion) and suggested CTA
Constraints: tone {{BRAND_VOICE}}; target {{AUDIENCE}}; focus on {{PRIMARY_GOAL}}.

Pro tip: Use subject lines with proven triggers (curiosity, urgency, benefit). Benchmarks as of 2026: average small-business open rates by email: 18–22%; subject lines that include a benefit and personalization can increase opens by 5–12% depending on list quality. Track opens and CTORs for continuous optimization.

Prompt 3 — SEO Content Brief (1,500–2,000 words)

Purpose: Generate a detailed SEO brief for a 1,500–2,000 word blog post that targets a priority keyword and supports SEM/organic growth.

System: You are an SEO content strategist with up-to-date SERP analysis as of July 2026. Provide structured JSON and an outline.

User:
Create an SEO brief for topic: "{{TARGET_KEYWORD}}". Provide:
- target_keyword: "{{TARGET_KEYWORD}}"
- search_intent: informational/commercial/navigational
- primary_target_audience
- word_count_recommendation: 1500-2000
- suggested_title (3 variants)
- meta_description (<= 160 chars)
- URL_slug
- headings_outline (H1-H3 structure with 6-10 sections)
- 8 LSI and long-tail keywords to include
- recommended internal links (2) and anchor text suggestions
- recommended external authoritative links (3) with reason
- top 3 competitor SERP URLs and short notes on why they rank
Constraints: Return JSON and a short writing brief (3-5 bullets) for the writer.

Actionable tip: Use the competitor URLs to run a content gap analysis using a tool like Ahrefs or SEMrush. If you connect ChatGPT-5.5 with your SEO tool via API, have it fetch top 10 SERP features and use them to target featured snippets (FAQ/definition sections). You can push this brief into your CMS or assign directly to a contractor.

Prompt 4 — Google Ads Responsive Search Ad Variations

Purpose: Generate 15 headlines and 4 description variants optimized for RSA experiments in Google Ads (character-constrained output).

System: You are a paid search copywriter for {{BUSINESS_NAME}}.

User:
Produce 15 concise headlines (<= 30 characters) and 4 description lines (<= 90 characters) for Google Ads Responsive Search Ads promoting "{{OFFER}}" targeting {{AUDIENCE}} in {{GEO}}. Include value props, urgency options, and a brand-safe call-to-action. Also provide 5 negative keyword suggestions.

Return as JSON:
- headlines: []
- descriptions: []
- negatives: []

Use case: Import these into Google Ads Editor or scripts for automated testing. Monitor CTR and conversion rate by headline group; rotate top performers into expanded text ads for control testing. Typical paid search KPI targets for SMBs (July 2026): CTR 3–8% depending on industry; conversion rate 2–12% depending on funnel and landing page quality.

Prompt 5 — Product/Service Landing Page Copy (Conversion-Optimized)

Purpose: Create a high-converting landing page with emotion-driven hero, benefits, proof elements, and a clear CTA. Include a short A/B test variant.

System: You are a landing page copywriter for conversion rate optimization.

User:
Write landing page copy for the product/service "{{PRODUCT_NAME}}" by {{BUSINESS_NAME}}. Provide:
- hero_headline
- subheadline
- 3 primary benefits (bullet list)
- 4 social proof elements (short testimonials or stats)
- pricing_summary (if applicable)
- CTA_text (primary & secondary)
- variant_B_changes (short list for A/B testing)
Return as HTML fragments only (no full page wrapper).

Note: Keep the hero headline focused on the primary benefit. If you have past conversion data, pass typical visitor behavior metrics in the variables for more tailored recommendations.

Prompt 6 — Content Repurposing Matrix

Purpose: Turn a single long-form asset into 12 repurposed items (social posts, micro-videos, newsletter snippets, SlideShare).

System: You are a content repurposing strategist.

User:
Given this content asset (title: "{{ASSET_TITLE}}", word_count: {{ASSET_WORD_COUNT}}, main_points: [list of 6 main points]), create 12 repurposed content items with:
- format (e.g., X thread, Instagram carousel, 60s TikTok script)
- exact copy or script
- suggested visual cue or slide structure
- recommended length
Return as a JSON array.

Pro tip: Repurpose the top-converting section as an email subject line test. For video scripts, keep the hook in the first 3 seconds; for carousels, map each slide to one micro-insight.

Prompt 7 — Local SEO GMB Optimization Checklist

Purpose: Produce an audit checklist for Google Business Profile (formerly GMB) and local citations tailored to your business category and location.

System: You are a local SEO specialist. Output a checklist in JSON and Markdown.

User:
Create a GMB/local SEO optimization checklist for {{BUSINESS_NAME}} in {{CITY, STATE}}. Include items for:
- primary category selection
- services list optimization
- photo strategy
- review acquisition and response templates
- citation audit (5 recommended local directories)
- local schema suggestions
Return both JSON and a short action plan with priority ranking (High/Medium/Low).

Actionable data: Businesses that complete profile optimizations and publish weekly posts to their Google Business Profile can see +15–25% local search visibility lift within 90 days in many categories (July 2026 industry averages).

Prompt 8 — Productized Service Packaging & Pricing

Purpose: Convert your hourly or bespoke service into three productized tiers (Basic, Pro, Premium) and produce positioning copy for each.

System: You are a pricing strategist for small businesses.

User:
Productize the service "{{SERVICE_DESCRIPTION}}". Provide:
- three tier names and price ranges (use local market median and +/- 20%)
- feature_list per tier (concise bullets)
- target_segment per tier
- positioning_blurb (1–2 sentences)
- recommended upsell/cross-sell
Constraints: Audience {{AUDIENCE}}, brand voice {{BRAND_VOICE}}.

Pricing note: Use local market research or SaaS benchmarking. If you provide past hourly rates, pass them as a variable to base the tiers on real margins.

Prompt 9 — Quarterly Marketing Plan (90-day roadmap)

Purpose: Produce a prioritized, measurable 90-day marketing plan aligning budget, channels, and KPIs.

System: You are a marketing director building a 90-day plan for {{BUSINESS_NAME}}.

User:
Create a 90-day marketing roadmap with:
- Objective (primary KPI)
- Budget breakdown by channel (SEM, social, content, email, partners)
- 12-week calendar with weekly milestones
- Expected outcomes and KPIs for each channel
- Two contingency/playbook actions if KPIs are below target after 4 weeks
Return as JSON and a 1-page summary paragraph.

Tip: For SMBs, allocate ~30–40% of marketing spend to direct response channels in early growth stages. Monitor CAC and LTV every 2 weeks.

Section 2 — Operations & Productivity (SOPs, Scheduling, Inventory)

Goal: Reduce manual work and operational risk with SOPs, scheduling assistants, and inventory automation. Recommended model settings: temperature 0.0–0.3 for deterministic process documents, max_tokens 1000–3000 depending on SOP complexity.

Prompt 10 — SOP Builder: Weekly Operations Playbook

Purpose: Create a 1–2 page Standard Operating Procedure (SOP) for a recurring operational task (daily opening, inventory restock, payroll run).

System: You are an operations manager building precise SOPs.

User:
Write an SOP for "{{TASK_NAME}}" to be followed by {{ROLE}}. Include:
- Objective (1 line)
- Frequency
- Estimated time required
- Prerequisites and tools (software, accounts)
- Step-by-step instructions (numbered, with expected time per step)
- Troubleshooting section (3 common issues with fixes)
- Required logs/outputs (file names or database fields)

Return as Markdown and a JSON checklist with actionable fields that can be used to build a checklist app.

Example task: "Daily bakery opening checklist" — include exact temperature checks, cash register reconciliation steps, and photo requirements for inventory. Use structured outputs to integrate with task managers like Asana or ClickUp via API.

Prompt 11 — Employee Shift Scheduler (Conflict-aware)

Purpose: Produce a weekly staff schedule given staff availability, roles, and labor-hour constraints.

System: You are a workforce scheduler.

User:
Create a weekly shift schedule for {{BUSINESS_NAME}} for week starting {{WEEK_START_DATE}}. Inputs:
- roles_needed_by_hour (JSON: hour -> roles & count)
- staff_list with availability, role(s), max_hours_per_week, hourly_rate, and seniority
- labor_budget_limit for the week

Constraints:
- No staff should exceed max_hours_per_week
- Required breaks must be scheduled (local labor laws)
- Optimize to minimize labor cost while meeting role coverage

Return as JSON with assignments per hour and a short explanation of trade-offs.

Implementation tip: You can chain this prompt with an integer programming solver in your automation stack for optimal schedules. For many SMBs, a greedy heuristic from ChatGPT-5.5 is sufficient and much faster to implement.

Prompt 12 — Inventory Reorder Triggers & Purchase Order Drafts

Purpose: Generate reorder triggers based on lead time, usage rate, and safety stock, and then draft a PO email/CSV for the supplier.

System: You are an inventory planner.

User:
Given SKUs with fields: sku_id, current_stock, avg_daily_usage (last 90d), supplier_lead_time_days, safety_stock_days, pack_size, unit_cost — calculate:
- reorder_point
- reorder_quantity (round to pack_size)
- suggested_po_date (considering lead time)
Return a CSV-ready table and draft a supplier email for the top 5 SKUs to reorder.

Return JSON and CSV table as code block.

Data-driven note: For many retail SMBs, setting safety_stock_days = 14 for seasonal items and 7 for staples reduces stockouts by ~35% compared to ad-hoc reordering (industry case studies as of 2025–2026).

Prompt 13 — Meeting Agenda & Follow-Up Automation

Purpose: Create a concise meeting agenda from key topics, produce a 30-minute time-boxed plan, and generate follow-up action items and email drafts.

System: You are an executive assistant.

User:
Create a 30-minute meeting agenda for meeting_title: "{{MEETING_TITLE}}". Inputs:
- participants and roles
- goals (3 max)
- pre-read materials (list of links)
Output:
- timeboxed agenda (with minute allocation)
- note-taking template (action_item, owner, due_date)
- follow-up email template that includes action items in bullet form

Return as JSON and an HTML email template.

Integration: Feed the note-taking template into your meeting recorder or use ChatGPT-5.5's transcription tools to auto-fill action items post-meeting for immediate assignment.

Prompt 14 — Vendor SLA & Contract Summary Extractor

Purpose: Provide a short, actionable summary of a vendor contract or Service Level Agreement (SLA) highlighting key dates, penalties, and auto-renewal clauses.

System: You are a contract analyst. You will use the document text provided.

User:
Summarize the attached SLA. Extract:
- effective_date
- renewal_terms and notice_period
- termination_conditions
- liability_limits
- response_time_and_escalation matrix
- payment terms
Return a one-page summary and a 5-point risk checklist.

If any auto-renew clause exists with less than 60 days notice, flag it as HIGH_RISK.

Actionable step: Use this prompt to triage contracts before signing or renewals. Many SMBs miss auto-renewal terms causing unexpected charges; a 60-day flagging rule prevents most surprises.

Prompt 15 — Project Kickoff Brief & RACI Matrix

Purpose: Generate a concise kickoff brief and a RACI matrix for small projects (2–6 week initiatives).

System: You are a project manager.

User:
Create a kickoff brief for project: "{{PROJECT_NAME}}". Inputs:
- scope_summary (3 sentences)
- key_deliverables (list)
- team_members and roles
- milestone_dates

Return:
- 1-page kickoff brief
- RACI matrix table
- 3 initial risks and mitigation steps

Use this to structure 2–6 week sprints and to quickly align stakeholders. A clear RACI reduces role confusion and speeds up approvals.

Section 3 — Finance & Accounting (Invoicing, Forecasting, Expense Tracking)

Goal: Reduce bookkeeping effort, improve cash flow forecasting, and standardize invoices and payment follow-ups. Recommended model settings: temperature 0.0–0.2 for precise formatting, max_tokens 800–1500 for multi-field outputs.

Prompt 16 — Invoice Draft & Overdue Payment Follow-up

Purpose: Generate a professional invoice and a sequence of three overdue payment follow-up emails tailored to customer behavior.

System: You are an accounts receivable specialist.

User:
Generate an invoice for:
- invoice_number: {{INVOICE_NUMBER}}
- bill_to: {{CLIENT_NAME}} (address, contact)
- items: [{description, qty, unit_price, tax_rate}]
- due_date: YYYY-MM-DD
Return a PDF-ready invoice HTML and a 3-step overdue email sequence:
- Day 1 (friendly reminder)
- Day 7 (stern, include late fee policy)
- Day 21 (last notice, include collection timeline)

Include payment links text and recommended subject lines. Return JSON and HTML for the invoice.

Best practice: Offer multiple payment methods. For small B2B invoices, providing ACH plus credit card options reduces days sales outstanding (DSO) by 7–12% on average for SMBs when offered at invoice time.

Prompt 17 — 12-Month Cash Flow Forecast (Monthly)

Purpose: Produce a 12-month cash flow forecast using known revenue streams, recurring expenses, and seasonal adjustments.

System: You are a financial analyst.

User:
Create a 12-month cash flow forecast for {{BUSINESS_NAME}} starting {{START_MONTH}}. Inputs:
- opening_cash_balance
- monthly_recurring_revenue (MRR) with growth assumptions (%)
- one-time_revenues (schedule)
- fixed_monthly_expenses (list)
- variable_expenses (% of revenue)
- tax_rate and expected one-time capital expenditures
Output:
- monthly cash_in, cash_out, net_cash_flow, ending_balance
- month-by-month cash runway estimate
- three scenario variants (base, best_case +15% revenue, worst_case -15% revenue)
Return as CSV and include short commentary on months at risk.

Actionable metric: Flag any month where ending_balance < 14 days of payroll as "At-Risk" and provide 3 prioritized actions (e.g., negotiate supplier terms, accelerate receivables, temporary staffing reduction).

Prompt 18 — Expense Categorization & Quarterly Tax Prep Notes

Purpose: Take a raw list of expenses and produce categorized buckets, estimated tax-deductible items, and preparation checklist for quarterly filings.

System: You are a small business bookkeeper.

User:
Given a list of expenses (date, description, amount, vendor), categorize each into: Cost of Goods Sold, Rent, Payroll, Subscriptions, Marketing, Travel, Professional Fees, Other. Identify potential tax-deductible items and flag transactions needing receipts.

Return:
- categorized_table (CSV)
- estimated quarterly_tax_liability (using provided revenue and tax_rate)
- checklist for document collection for quarterly filing
Return JSON and CSV.

Tip: For many SMBs, consistent categorization reduces accountant fees during tax season by 20–40% and prevents missed deductions.

Prompt 19 — Unit Economics & Pricing Sensitivity Analysis

Purpose: Compute unit economics (unit contribution margin, CAC payback, LTV:CAC) and produce price sensitivity scenarios to guide pricing decisions.

System: You are a unit economics analyst.

User:
Calculate unit economics for product "{{PRODUCT_NAME}}". Inputs:
- avg_price
- variable_cost_per_unit
- gross_margin_pct
- avg_CAC
- customer_lifetime_months
- monthly_churn_rate
Output:
- unit_contribution_margin
- payback_period_months on CAC
- LTV (and LTV:CAC ratio)
- 3 pricing scenarios (+10%, -10%, +25%) with effect on margin and payback
Return as JSON and a short recommendation.

Benchmarks: For subscription SMBs in 2026, aim for LTV:CAC >= 3 for sustainable growth; for one-time product sales, target contribution margins > 40% unless volume economics compensate.

Section 4 — Customer Service (Response Templates, FAQ Generation, Complaint Handling)

Goal: Improve CSAT, reduce response time, and automate common reply patterns. Recommended model settings: temperature 0.1–0.3 for consistent templates; use function calls for ticket updates and CRM writes.

Prompt 20 — AI-Powered FAQ Generator (From Transcripts or Docs)

Purpose: Generate an FAQ page from a product manual or recorded customer support transcripts.

System: You are a knowledge base engineer.

User:
Given input text (product_manual_or_transcripts), extract the top 30 FAQs, provide a short answer (1-2 sentences), required follow-up actions (if any), and categorize by topic. For each FAQ, produce:
- question
- concise_answer
- related_article_link_suggestion (slug)
- priority_rating (High/Medium/Low)
Return JSON and a ready-to-paste Markdown FAQ section for the website.

Implementation: Import the JSON into your knowledge base platform (Zendesk Guide, Intercom Articles) and map categories to article tags. Use analytics to monitor which FAQs reduce ticket volume.

Prompt 21 — Customer Support Response Templates (Tone-Checked)

Purpose: Provide templates for common inbound inquiries with variable placeholders and escalation guidance.

System: You are a customer support lead enforcing brand voice.

User:
Create templates for these ticket types:
1) Refund request within 30 days
2) Technical issue (app crash)
3) Billing dispute
4) Order status inquiry (domestic)
For each template include:
- subject_line
- opening_paragraph (acknowledge & empathize)
- troubleshooting steps or resolution options
- escalation instructions (when to escalate to level 2)
- placeholders to personalize (e.g., {{ORDER_ID}})
Return JSON and plain text versions ready for copy-paste.

Note: Personalization increases CSAT by 7–10% when template placeholders are reliably filled. Use CRM merge fields where possible to populate variables automatically.

Prompt 22 — Complaint Handling Playbook (De-escalation Scripts)

Purpose: Provide a short playbook with de-escalation scripts, refund/credit rules, and a 5-step process to close complaints within 48 hours.

System: You are a customer experience specialist.

User:
Draft a 5-step complaint handling playbook for complaints that are likely to escalate on social channels. Include:
- immediate response scripts (first 24 hours)
- criteria for issuing goodwill credits or refunds (monetary thresholds)
- public response guidelines (what to post publicly vs. DM)
- escalation matrix with response SLAs
Return bulletized playbook and three example public responses for typical complaints.

Evidence: Businesses that resolve complaints publicly within 24–48 hours often see an increase in brand perception and can reduce churn; quick public resolutions can reduce follow-on social amplification by ~60%.

Prompt 23 — Chatbot Handover & Human Escalation Rules

Purpose: Generate a rule set that defines when an automated bot response should escalate to a human agent, including sample handover messages to preserve context.

System: You are an AI chatbot designer.

User:
Create escalation rules for a customer service chatbot for {{BUSINESS_NAME}}. Input:
- complexity_threshold (e.g., user uses words: "cancel", "law", "refund", "chargeback")
- sentiment_threshold for escalation (e.g., low sentiment or explicit "angry")
- max_bot_turns (e.g., 3)
Output:
- structured rule set (JSON) of triggers
- sample human-handover message including last 3 user exchanges
- suggested wait-time SLA for human response after escalation
Return JSON and short implementation notes.

Tip: Include user transcript excerpts for context during handover so humans can act without re-asking obvious questions — reduces average handle time by 20–30%.

Prompt 24 — CSAT Survey with Branching Logic

Purpose: Create a 6-question CSAT survey with conditional follow-ups that can be sent after ticket closure.

System: You are a customer insights manager.

User:
Design a post-interaction CSAT survey of max 6 questions. Include:
- a 1–5 CSAT numeric question
- conditional questions if score <= 3 to capture issue category and permission to follow up
- NPS style question if score >= 4
- suggested timing (immediate, 24h, 72h) and a subject line for the email/SMS
Return JSON with branching logic and CSV-ready header row for responses.

Best practice: Send the primary CSAT query within 24 hours of resolution. Short surveys increase response rates; aim for 8–12% response rates for email surveys and higher for in-app surveys.

Prompt 25 — Refund & Returns Policy Draft Generator

Purpose: Create a clear, legally considerate returns and refunds policy for e-commerce SMBs with dynamic clauses for product type and shipping.

System: You are a commercial policy writer.

User:
Draft a returns and refunds policy for {{BUSINESS_NAME}} selling {{PRODUCT_CATEGORY}} with:
- standard return window (e.g., 30 days)
- process steps for customer (return label, restocking fee)
- exceptions (final sale, perishables)
- sample customer-facing FAQ blurb
Return as plain text and a short version for the product page (<= 150 words).

Note: This is not legal advice. Flag clauses that should be reviewed by a lawyer.

Operational suggestion: Put a copy of this policy into your order confirmation emails and checkout pages. Clear policy placement reduces returns-related inquiries by ~12% on average.

Prompt Engineering Best Practices & Model Settings

Below are practical, actionable guidelines for deploying these prompts effectively with ChatGPT-5.5.

System & User Message Strategy

• Always send a brief system prompt with constraints: brand voice, output format, and forbidden actions (e.g., "Do not include any external links unless provided").
• Pass contextual variables in the user message but keep them explicit and labeled. Use a JSON payload if your client supports it.

Temperature, max_tokens, and Determinism

• Copy/structured outputs: temperature 0.0–0.2 (deterministic)
• Creative brainstorms (campaign slogans, A/B variants): temperature 0.4–0.7
• Long procedural documents: temperature 0.0–0.3 and enable "streaming" if supported

Structured Output Formats

When integrating with downstream systems, request JSON or CSV explicitly. For example: "Return strictly valid JSON. If you cannot provide JSON, return a JSON with an 'error' field." This avoids extraneous commentary that breaks parsers.

Versioning & Auditability

• Embed a prompt version_id and timestamp in all generated artifacts so you can audit changes (e.g., "prompt_version": "v1.3", "generated_at": "2026-07-15T10:00:00Z").
• Store outputs with input variables in your DB for reproducibility and compliance.

Testing & Rollout

Test each prompt in a staging environment with sample inputs and edge-case values (nulls, extreme numbers, multilingual text). Run A/B tests for content and pricing prompts; measure lift against existing baselines before permanent adoption.

Summary Table: Quick reference of the 25 prompts

Integration Tips: APIs, Zapier, Google Sheets, and Function Calls

This section provides practical recipes for integrating these prompts into automated workflows for small teams.

Recipe A — Weekly Social Calendar to Scheduling Tool (Zapier/GPT-5.5)

1. Run Prompt 1 on Monday morning via the API or ChatGPT interface.
2. Get JSON output and validate schema (date, platform, post_copy, hashtags, image_desc).
3. Use Zapier or Make to create scheduled posts in your scheduler (Buffer/Meta Business Suite) by mapping fields. If image assets are described but not available, trigger a task for the design team (Asana/Trello ticket) with the suggested image description.

Recipe B — Invoice Generation to Accounting (Google Sheets + Stripe)

1. Trigger Prompt 16 when an order is fulfilled.
2. Store the HTML invoice in Google Drive and generate a public (or signed) URL.
3. Push invoice metadata to Stripe via API or send the invoice link to the customer via transactional email (SendGrid/Postmark).

Function Calls & Tool Use

If your ChatGPT-5.5 instance supports function calling, define functions for "create_calendar_event", "create_invoice", "send_support_email". Return structured function arguments in the prompt and let the platform execute them. This reduces the need for intermediary parsing: the model returns function call payloads ready for execution.

Further Reading & Internal Resources

For deeper implementation guides, connectors, and example projects, see these internal resources: AI prompts entrepreneurs, ChatGPT for SMBs, and business automation prompts.

Final Notes & Governance

Operationalize with care: maintain human oversight for high-risk outputs (legal language, large payments, sensitive customer interactions). Log all inputs and outputs, version your prompts, and schedule prompt audits quarterly to adapt to market and model updates. In July 2026, AI model capabilities and platform integrations continue to evolve — keep developer and vendor release notes on your watchlist and test automated flows after any model update.

Use this playbook as the blueprint for immediate efficiency gains. Start with 3–5 prompts that solve your highest time-sink tasks, measure impact (time saved, ticket reduction, revenue uplift), then expand automation across the organization.