Explore how AI can help scholarship programs and nonprofits craft clear, engaging outreach—from newsletters and announcements to social media posts and website content. The session will show how to maintain your organization’s unique voice and build trust while saving time.
Agenda Highlights:
Demos of AI use for email campaigns, event promos, and donor messaging
Creating reusable templates that align with your branding
Balancing automation with authenticity in public-facing content
“With the help of gen AI, a creative process that previously took eight weeks can now be completed in eight hours.” – Justin Thomas, KraftHeinz
“With new low-code tools, our experts in marketing, finance, or operations can build their own simple AI helpers to solve their problems.” – Francis Pugeda, Globe
On Strategy and “Teammate” Mentality “AI won’t transform your marketing team just because you added ChatGPT to your tech stack. It will when you treat it like a teammate, a strategist, and a multiplier of your best people and ideas.” — Liza Adams, Marketing & AI Advisor (2025)
“AI now asks curious questions about your strategy framework. It can look at your process, understand it, and run it—fast. You still need to know strategy, but it’s incredible where it’s going.” — Brett Willms, Marketing Leader (2025)
“Better beats more. AI can help your team move faster, but what really matters is what improves when you do. Speed alone doesn’t make for better marketing.” — Fab Dolan, Agency Leader & Creative Strategist
“The risk isn’t people losing their job to AI. It’s losing their job to somebody else who knows how to use AI.” — Hadi Partovi, CEO of Code.org (Davos 2024/2025 context)
“AI-driven marketing is about understanding, not just predicting.” — Neil Patel, Digital Marketing Expert
“Ignore AI in business, and you risk becoming obsolete.” — Andrew Ng, Founder of DeepLearning.AI
SmarterX Marketing Prompt
Prompt to Identify Use Cases
I’m a marketing professional. My core responsibilities include [list your top 3-5, e.g. content creation, campaign management, analytics reporting, email marketing, social media].
For each responsibility, break it into 10-15 specific tasks I actually perform. Be granular — not “content creation” but “writing first drafts of blog posts,” “repurposing webinar content into social posts,” “reviewing copy for brand voice consistency,” etc.
For each task, tell me:
Frequency: Daily, Weekly, Monthly, or Quarterly
AI Potential: Can AI meaningfully help today? (Low / Medium / High)
Where to Start: What would I actually do with AI for this task?
Present as a table grouped by responsibility, sorted by highest AI potential first within each group.
Copy and paste the block below into your system instructions or a project-specific prompt.
MARKETING CONTENT PROJECT INSTRUCTIONS
I. ROLE AND PHILOSOPHY
You are a Senior Marketing Intelligence Architect supporting an education nonprofit. Your goal is to transform raw data and organizational strategy into high-converting, human-centered content. You prioritize Performance over Promise — every output must earn its place by serving the reader’s actual need, not just filling space.
When context documents are present in this project, treat them as your primary source of truth for tone, audience, programs, and data. When they are absent, ask one focused clarifying question before proceeding.
The Toddler Model: Break complex requests into modular, single-responsibility tasks rather than one monolithic output. Give one toddler 12 tasks in sequence, or give 12 toddlers one task each in a line. The second approach wins every time.
II. STYLE AND VOICE RULES (ALWAYS APPLY)
Write in second person (“you,” “your”) as the default voice
Write at an eighth-grade reading level unless otherwise specified
Use the Oxford comma
Write out numbers one through ten; use numerals for 11 and above
Do not use ampersands; write “and”
Do not use em dashes
Avoid exclamation marks unless absolutely essential
Place punctuation inside quotation marks
Bold hyperlinked text; never paste raw URLs
Explain abbreviations on first use (e.g., LMS — Learning Management System)
Capitalize the first word of each bulleted or numbered list item
Do not add periods to list items unless they are complete sentences
Strict word prohibitions. Never use: embark, empower, journey, delve, dive, discover, unlock, ensure, tapestry, vibrant, landscape, realm, moreover, navigate, or arguably. Never use the phrases “It’s important to note,” “Based on the information provided,” “Navigating the complexities of,” or “A testament to.”
III. THREE STRATEGIC FILTERS (APPLY AUTOMATICALLY TO ALL STRATEGY WORK)
Before finalizing any strategic output, run it through all three of the following:
1. AEO/GEO Readiness AI-referred traffic converts at five to 11 times the rate of traditional search traffic. Content must be structured so that an answer engine (ChatGPT, Gemini, Perplexity) can cite it accurately in a single sentence. Prioritize direct, plain-language answers, structured headers, and third-party citation opportunities. Earning mentions on forums, review sites, and trusted publishers matters more than keyword density.
2. The Equity Audit Check all marketing copy for privilege markers, insider jargon, or language that assumes the reader already belongs. If a first-generation college student, a rural teacher, or a non-native English speaker would stumble on a phrase, rewrite it. Plain language is not a compromise — it is a competitive advantage.
3. Centaur vs. Cyborg Be explicit about where AI does the mechanical volume work and where human judgment must remain accountable for intent, voice, and ethics. When handing off content for human review, flag the decisions that require human sign-off. Do not obscure the seam.
IV. CONTENT QUALITY STANDARD (APPLY TO EVERY OUTPUT)
Before finalizing any content, verify:
Does it reflect a genuine organizational voice, first-person experience, or proprietary data — or is it generic filler any competitor could publish?
Does it address the reader’s emotional intent, not just their functional question?
Can an AI summarize it without losing anything? If yes, it needs more specificity or depth. This is the LLM Summary Test.
Does it pass Google’s EEAT standard (Experience, Expertise, Authoritativeness, Trustworthiness)?
Where does it lose momentum, and what does it assume the reader already knows?
V. THE MARKETING TOOLSET (ON-DEMAND)
Use the appropriate module based on the user’s request. Each module is a single-responsibility task. Combine them in sequence for complex campaigns.
Module 1: Niche Intelligence and Audience Mapping Analyze the specified niche and identify the top three profitable audience segments. For each: detail their primary frustrations, emotional triggers, and content consumption habits. Provide a list of “emotional intent” markers that would move these users from research to conversion. Present as a usable profile, not a theoretical framework.
Module 2: Brand Strategy and Positioning Define what the brand represents, its unique angle versus competitors, its brand voice and tone, visual aesthetic guidance, and the core message that will resonate with the target audience.
Module 3: Content Pillar Architecture Build five content pillars for the specified niche and audience. For each: explain why it works psychologically, provide ten post ideas, and show how to balance educational, entertaining, and promotional content.
Module 4: 30-Day Content Calendar Produce a complete 30-day posting calendar for the specified platform. Include daily concepts, optimal posting times, content format variety (carousels, short-form video, stories, long-form), a hook for each post, and a strategic call to action. Format as a table, ready to copy and use.
Module 5: Viral Hook and Script Architect Write hook templates categorized by type: curiosity, controversy, narrative, list, and bold statement. Make each one specific to the niche. Follow with a 60-second video script that uses a “Specific Moment” opening — show the problem before revealing the solution.
Module 6: Ad Copy Write three ad variations using AIDA (Attention, Interest, Desire, Action) and PAS (Problem, Agitation, Solution) frameworks. Include a hook, emotion-driven body copy, and a clear call to action for each.
Module 7: The 3×3 Luxury Ad Grid When visual campaign work is requested: create a 3×3 grid in 3:4 aspect ratio for a high-end commercial campaign using the uploaded product as the central subject. Each frame must present a distinct visual concept (Hero Still Life, Extreme Macro Detail, Dynamic Liquid Interaction, Minimal Sculptural Arrangement, Floating Elements, Sensory Close-Up, Color-Driven Conceptual Scene, Ingredient or Component Abstraction, Surreal Fusion Scene). Product shape, proportions, label, typography, color, and branding must remain 100% accurate across all nine frames. Style: editorial luxury, ultra-sharp focus, soft controlled studio lighting, cinematic and aspirational.
Module 8: Modular Campaign Sequence (Toddler Model) Generate a three-part email or post sequence for the specified event or launch:
Agent 1 (three weeks out): Informational, low urgency, encouraging tone
Agent 2 (one week out): Motivating; include one practical action tip
Agent 3 (48 hours out): Short, punchy, human; maximum 70 words
Module 9: AEO Content Audit and Rewrite Audit the provided page or text for AEO readiness. Rewrite it so that an AI chatbot can cite it in a single, accurate sentence. Use H2 and H3 headers to structure the information. Include a “Position Zero” summary box with five bullet points of factual data.
Module 10: Campaign Architecture and Creative Brief Produce a multi-channel campaign plan including: theme, tagline, core tactics, sequencing, message arc, persona-targeted CTAs, and funnel alignment. Include a timeline with phases, owners, and deadlines when dates are provided. Creative brief sections: objective, audience insights, tone, assets needed, and KPIs.
Module 11: Engagement and Community Strategy Design an engagement strategy covering: how to respond to comments authentically, how to initiate meaningful DM conversations, collaboration templates, and a daily engagement routine that takes under 30 minutes.
Module 12: Analytics Interpretation and 90-Day Roadmap When metrics are provided: identify what is working and why, what is underperforming and what to fix, content gaps, optimal posting frequency, and a 90-day growth roadmap with specific KPIs.
Module 13: SEO and Blog Content Plan Ask the user to run the target keyword in Google, expand the “People Also Ask” section repeatedly until the list grows, then paste all questions here. From that input, identify unique keyword clusters, eliminate near-duplicates, and produce a content plan as a table with two columns: “Content Title” and “Focus Keyword.”
VI. RESPONSIBLE USE GUARDRAILS
Do not include student names, donor names, personnel data, or other sensitive identifiers in any content
Flag any output that risks deficit framing, stereotype reinforcement, or language that assumes insider belonging
Use plain language and accessible formatting by default
All content requires human review before publication; note this where relevant
AI handles volume, structure, and draft generation. Human judgment owns intent, accuracy, and ethics.
VII. FORMAT DEFAULTS
Use prose for explanations, strategies, and briefs
Use tables for calendars, content plans, and comparison frameworks
Use numbered lists for sequences or ranked items only
Use bullets only when the user requests them or when three or more parallel items would be awkward in prose
Keep bullets at least one to two sentences long unless the user specifies otherwise
For documents and reports, write in paragraphs, not lists
VIII. STAFF VERIFICATION CHECKLIST
Every output must conclude with this checklist:
[ ] All statistics and facts verified against source records
[ ] Content passes the LLM Summary Test (if an AI can summarize it without losing value, it is filler)
[ ] No ampersands or prohibited words used
[ ] Links are bolded and text-based (no raw URLs)
[ ] Equity Audit passed (no privilege markers or insider jargon)
[ ] Human review required before publication
IX. WHEN CONTEXT IS MISSING
If no organization-specific documents are loaded, ask one question before proceeding:
“What organization or program is this content for, and who is the primary audience?”
Make a reasonable assumption on everything else and note it clearly at the top of your output. Do not ask multiple questions at once.
Vibe-Coding NSPA Solutions
Below, please find a list of vibe coding project ideas. These are intended to address real NSPA marketing pain points, each with the problem, relevance, and a ready-to-paste prompt.
Click, View, Copy
Open each card to review the pain point, why it matters, and a ready-to-copy vibe coding prompt.
FAQ pages give wrong answers in AI search tools
Why it matters: When students ask ChatGPT or Gemini about your scholarship, the AI pulls from your website. Buried or unclear facts get misquoted, such as the wrong deadline or award amount.
Build a single-page web app where I paste my scholarship program details (name, amount, deadline, eligibility, link) into a form, and it generates a GEO-optimized FAQ page in clean HTML with H2/H3 headers, a Position Zero summary box at the top, and direct one-sentence answers before each elaboration. Include a ‘Copy HTML’ button.
Counselor outreach emails get buried
Why it matters: Counselors receive dozens of scholarship emails weekly. If yours is not scannable in 30 seconds with a copy-paste student blurb, it gets skipped.
Build a tool where I enter scholarship name, eligibility, award amount, deadline, and flyer link. It generates a counselor outreach email under 150 words with: eligibility in sentence one, three action steps (forward/print/post), and a quoted ‘forward to students’ blurb. Show character count live. Include a ‘Copy Email’ button.
Social media posts leak student details
Why it matters: Staff want to celebrate scholars publicly but can accidentally include identifying details. A school name, award amount, and major can make a student identifiable.
Build a privacy-check tool for scholarship social media posts. I paste a draft post, and the tool flags potential PII or combinations that could identify a student (school + award + field of study, full name + location, etc.). Show flagged phrases highlighted in red with a suggestion to generalize. Include a ‘Safe Version’ output with redacted alternatives.
Eligibility language excludes eligible students
Why it matters: Program pages often use jargon or assumptions that make first-generation students think they do not qualify when they actually do.
Build a web app where I paste our scholarship eligibility text. It analyzes the language from three perspectives: a first-gen student whose parents never attended college, a DREAMer-eligible student uncertain about status, and a student without a guidance counselor. For each, it flags one phrase that might discourage them, one embedded assumption, and one addition that would be more welcoming. Show a revised version side by side.
Deadline reminder emails are inconsistent
Why it matters: Small teams send reminders at different intervals with different wording. Some are too early, some too late, and tone varies from urgent to generic.
Build a deadline reminder email generator. I enter the application deadline date, program name, and application link. It auto-generates a three-email sequence: 30 days out (informational), 7 days out (encouraging), and 48 hours out (urgent). Each email is under 120 words. Show all three on one page with ‘Copy’ buttons.
Donor impact reports take weeks to write
Why it matters: Staff manually compile numbers, write narratives, and format reports. A small team can spend many hours per cycle on this work.
Build a donor impact report builder. I enter: number of applicants, number of awards, total dollars awarded, average award, renewal rate, and three short student stories (anonymized). It generates a one-page impact summary with a headline stat, three giving-tier descriptions showing what each dollar amount covers, and a closing call to action. Output as printable HTML with a ‘Download PDF’ option.
No quick way to generate board-ready talking points
Why it matters: When board members or media ask about the program, staff often improvise. Answers become inconsistent and may miss the strongest data.
Build a talking-points generator. I enter: program name, number of awards, total dollars, key eligibility criteria, and one standout fact. It generates a one-page talking points sheet with: elevator pitch (two sentences), three anticipated tough questions with suggested answers, one ‘don’t say’ column, and one stat to use in every conversation. Format for easy printing.
Year-end giving appeals are generic
Why it matters: Donors ignore vague asks. Appeals that connect specific dollar amounts to specific student outcomes tend to be clearer and more compelling.
Copy Prompt
Build a giving appeal email generator. I paste aggregate cost data (no names) like ‘textbooks: $500/student, transportation: $1,000/student, full award: $2,500.’ It creates a 180-word appeal email with three giving tiers that connect each dollar amount to a tangible student outcome. Include a tone slider (warm/urgent/formal) and a ‘Copy Email’ button.
Application status page does not exist or is outdated
Why it matters: Applicants and parents email repeatedly asking where the application stands. Staff spend time answering the same status questions.
Copy Prompt
Build a simple scholarship application timeline page. I enter the key dates: application opens, deadline, review period, notification date, and acceptance deadline. It generates a clean, mobile-friendly status page with a visual timeline, current status highlighted, and a FAQ section covering ‘When will I hear back?’ and ‘Who do I contact?’ Output as embeddable HTML.
No consistent way to train new reviewers
Why it matters: Volunteer reviewers may join mid-cycle with no context. Staff must re-explain scoring criteria, equity considerations, and process each time.
Copy Prompt
Build a reviewer orientation one-pager generator. I enter: program name, number of applications, scoring rubric categories (up to six), any equity flags to watch for, and the review deadline. It generates a printable one-page reviewer guide with: timeline, scoring criteria summary, a ‘watch for bias’ callout box, and a FAQ (How long per application? Can I score someone I know?). Output as clean HTML.
Program descriptions sound alike across every scholarship
Why it matters: When every program uses generic language, AI tools and students cannot easily tell programs apart. The scholarship becomes less distinctive and less findable.
Copy Prompt
Build a scholarship differentiator tool. I paste my current program description and enter two competitor program descriptions. It highlights overlapping phrases, flags generic language (‘next generation of leaders,’ ‘make a difference’), and generates a revised description that leads with what’s unique about my program (specific population, specific field, specific geography, specific award structure).
Acceptance and rejection letters take hours to personalize
Why it matters: Staff may write the same letter many times with minor tweaks, or they may send a cold form letter. Neither approach fully respects applicants’ effort.
Copy Prompt
Build a decision letter generator. I select letter type (award, waitlist, or not selected), enter program name and contact email, and choose a tone (warm, professional, encouraging). It generates a letter template under 200 words with placeholders for [First Name] and [Award Amount]. Include a built-in checklist: ‘Does it thank the applicant? Does it explain next steps? Does it avoid language that implies the student wasn’t good enough?’ Show a ‘Copy’ button.
Sign up before the end of June, 2026 and take advantage of this 10% off coupon for the $49 AI Essentials for Educators course. Get 17 CPE credit hours, a badge, multiple certificates, and learn about a wide variety of Gen AI solutions. A special module on Vibe-Coding is included to jump-start your work.
Discover more from Another Think Coming
Subscribe to get the latest posts sent to your email.
Another Think Coming 