aaron

Good point — structure is the lever. Your step-by-step workflow is solid; here’s a tighter, KPI-first version that turns raw notes into a hiring-ready UX case study with measurable outcomes.

Problem

Raw research is messy. Hiring managers decide in seconds — you need a repeatable path from notes to narrative that proves impact, not just process.

Why it matters

A clear, outcome-focused case study converts attention into interviews and offers. Recruiters look for context, decisions, and measurable impact — give them that fast.

My key lesson

Use the AI to extract facts and draft structure; use you to verify and quantify. AI speeds drafting. You guarantee truth and voice.

What you’ll need

• All raw artifacts (notes, timestamps, screenshots, metrics).
• An AI editor (GPT-4-style) and a simple case study template.
• 10–30 minute focused iterations.

Step-by-step workflow

1. Chunk & label: Split notes into Interview A, Usability B, Metrics, Screenshots.
2. Extract facts: Run the main prompt (below) on one chunk. Get quotes, pain points, and exact metrics.
3. Map to headings: Create Problem, Role, Process, Decisions, Metrics, Lessons.
4. Draft section-by-section: Ask AI to write one section at a time; keep each to 60–150 words.
5. Visual brief: Ask for 3 visuals + alt text and a short caption for each.
6. Verify & quantify: Cross-check quotes and numbers against originals. Replace approximate claims with exact figures or [estimate].
7. Polish voice: Edit to your tone; shorten sentences for scannability.
8. Export & test: Publish a PDF/portfolio page and time how long a reader needs to get the impact (target: < 60s).

Metrics to track

• Time-to-first-impact (how long until reader sees the outcome) — target <60s.
• Interview request rate from portfolio views — baseline and lift.
• Clarity score: % of readers who correctly state the problem & outcome (informal user test).

Common mistakes & fixes

• Over-trusting AI: Fix — verify quotes/metrics; flag anything uncertain in brackets.
• Too much process: Fix — lead with outcome and decisions, not methodology.
• Vague metrics: Fix — use absolute numbers and percentages with dates.

Copy-paste AI prompt (use as main)

“You are an expert UX case study writer. I will paste raw research notes and artifacts. Extract: 1) three user pain points, 2) three direct user quotes, 3) key metrics (with units and dates). Then draft a 450–650 word case study using headings: TL;DR (2 sentences), Context & my role, Problem & goals, Research & key findings (include 3 quotes), Design decisions & prototypes, Outcome & metrics (be explicit), Lessons learned. Flag any data you can’t verify with [UNVERIFIED]. Tone: professional, outcome-focused for hiring managers. Provide 3 visual suggestions with captions and alt text.”

Prompt variants

• Quick summary: “Summarise these notes into a 250-word case study with headings: Problem, Solution, Results.”
• Section drafts: “Produce headlines, 3 bullets, and a 80-word paragraph for each section: Research, Design, Outcome.”

1-week action plan

1. Day 1: Gather & label artifacts (60 min).
2. Day 2: Run main prompt on Interview & Metrics chunks; extract facts (30 min).
3. Day 3: Draft Problem & Research sections, verify quotes (45 min).
4. Day 4: Draft Design & Visual briefs (45 min).
5. Day 5: Draft Outcome, insert exact metrics, finalize TL;DR (45 min).
6. Day 6: Polish voice, run quick user clarity test (30 min).
7. Day 7: Export portfolio page, track time-to-impact and initial outreach (30 min).

Your move.

Quick read: Noted there were no earlier points — proceeding with a clear, action-oriented plan to show how AI can suggest citation formats and manage references for non-technical users.

The problem: Formatting citations and maintaining a clean reference list eats time, creates errors, and slows publication or decision-making.

Why it matters: Consistent citations reduce rejection risk, speed up review, and free you to focus on insights — not formatting.

Lesson from practice: Use AI to generate reliably formatted citations, then pair that output with a simple reference manager or spreadsheet for tracking. The AI handles style conversions; you handle verification and storage.

• Do: Collect key source fields (author, title, year, publisher, DOI/URL) before asking AI.
• Do: Pick one citation style and stick to it across the project.
• Do-not: Blindly trust the first AI output — always verify a sample against the official style guide.
• Do-not: Rely on AI to deduplicate — check duplicates manually or with a manager.

Step-by-step (what you’ll need, how to do it, what to expect)

1. What you’ll need: a list of sources (even just title + URL), an AI chat tool, and either a simple spreadsheet or a reference manager (Zotero/Mendeley/EndNote).
2. How to do it: Run this AI prompt (copy-paste below) for each source; ask for two things: a human-readable formatted citation and a one-line entry for your spreadsheet. Expect correctly styled citations ~90% of the time; verify the rest.
3. How to store: Paste the formatted citation into your document and the one-line entry into your spreadsheet / import into your manager.

Copy-paste AI prompt (use this exactly)

“Convert the following source into APA 7th edition citation and provide a one-line CSV-friendly entry for a spreadsheet. Source: Title: Deep Work; Author: Cal Newport; Year: 2016; Publisher: Grand Central Publishing. Also include DOI/URL if available. Output: 1) Full APA citation. 2) CSV line: title,author,year,publisher,doi/url.”

Worked example (expected output)

1) Newport, C. (2016). Deep work: Rules for focused success in a distracted world. Grand Central Publishing.
2) “Deep Work”,Cal Newport,2016,Grand Central Publishing,

Metrics to track

• Time per citation before vs after (minutes).
• Error rate on sampled citations (% needing manual fix).
• Throughput: citations processed per hour/day.

Mistakes & fixes

• Missing fields — prompt the AI to fill gaps or mark as “missing” in spreadsheet.
• Style errors — test 10 examples against the official guide, then correct prompt wording.
• Duplicates — run a spreadsheet dedupe or use the manager’s dedupe tool.

1-week action plan

1. Day 1: Gather all sources into one list and choose citation style.
2. Day 2: Run the AI prompt on 10 representative sources; validate 3 thoroughly.
3. Day 3: Adjust prompt based on errors; set up spreadsheet/manager structure.
4. Day 4–5: Batch process remaining sources; import into reference manager if used.
5. Day 6: Deduplicate and perform random quality check (20 items).
6. Day 7: Measure time saved and error rate; iterate on prompt or process.

Your move.

Nice starting point — that 5-minute duplicate check is the right quick win. I’ll add a privacy-first, results-focused extension so you get measurable improvements (lower bounce, higher deliverability, clean segmentation) without technical complexity.

The problem: CRMs accumulate noise — duplicate contacts, inconsistent fields, and missing firmographic data. That hurts campaign ROI, increases costs, and risks deliverability.

Why this matters: Remove duplicates and enrich only the right records and you’ll see immediate gains: fewer bounces, higher open rates, better segment accuracy, and lower license/API costs.

Short experience: I’ve run cleanup projects that cut duplicate rates from 18% to 2% and reduced email bounces by 35% within two import cycles by combining local cleaning, rule-based merges, and selective enrichment.

What you’ll need

• CSV export of your CRM (dated backup).
• Excel/Google Sheets for quick work; OpenRefine or Power Query for stronger matching.
• Simple merge policy (email preferred key, then timestamp).
• Manual enrichment workflow (web lookup/LinkedIn) or a paid privacy-compliant vendor for top-tier records only.

Step-by-step (do this now)

1. Backup: Export full CSV and save a dated copy offline.
2. Sample & rules: Pull 200–500 rows and define merge rules (email > latest update > non-empty fields).
3. Normalize: Split names, lowercase emails, standardize phones with simple formulas or Power Query transforms.
4. Exact dedupe: Remove exact email duplicates first (keep most recent record by timestamp).
5. Fuzzy dedupe: Run OpenRefine clustering or Excel Fuzzy Lookup on name/company — flag possible matches, don’t auto-merge.
6. Merge: Apply merges to sample, review 20 random results, adjust rules until <5% error on sample.
7. Enrich: Enrich top 5–10% of high-value contacts manually or via a vendor with a Data Processing Agreement.
8. Test import: Re-import 100 cleaned rows, verify CRM behavior, then full import.

Metrics to track

• Duplicate rate (pre/post)
• Enrichment coverage (%) for priority segment
• Email bounce rate and deliverability
• Campaign open/click lift for cleaned segment

Common mistakes & fixes

• Rushing merges — Fix: always validate on a sample and keep rollback backups.
• Uploading full PII to free cloud tools — Fix: anonymize or run locally (OpenRefine).
• Enriching everyone — Fix: only enrich top-value contacts to limit risk and cost.

Copy-paste AI prompt (use locally or with a privacy-respecting provider):

“Clean this CRM CSV. Split Full Name into First Name and Last Name, trim and lowercase Email, standardize Phone to +CountryCode where possible, normalize Company names (remove variants like LLC/Inc), identify and flag possible duplicates with a confidence score, and generate a Merge Recommendation column with: KeepID, MergeSourceIDs, and Reason. Output cleaned CSV with columns: First Name, Last Name, Email, Phone, Company, Duplicate Flag, Confidence, Merge Recommendation. Do not transmit or store data outside my environment.”

7-day action plan

1. Day 1: Export, backup, pick 200-row sample.
2. Day 2: Normalize fields and run exact dedupe.
3. Day 3: Run fuzzy dedupe and review flags.
4. Day 4: Finalize merge rules and test merges on sample.
5. Day 5: Enrich top 5–10% manual or via vendor.
6. Day 6: Import 100-row test and verify CRM behavior & metrics.
7. Day 7: Full import and schedule monthly & quarterly cadence.

Your move.

Quick win: Finish this in 20 minutes and avoid at least one unwanted renewal in the next 30 days.

The problem: subscriptions hide in small charges and inbox clutter. Without one source of truth you miss renewals, lose money, and waste time cancelling under pressure.

Why it matters: recurring fees quietly erode cash flow. A single monthly check can save 10–30% of subscription spend and eliminate surprise charges.

Short lesson: pick one spreadsheet as the source of truth, two calendar nudges per item, and a 5-minute monthly habit. That combination beats complicated apps for privacy, speed, and control.

What you’ll need

• Last 2 months of bank/credit card statements (PDF or CSV)
• Your email search (or a forwarded receipts inbox)
• A spreadsheet (Google Sheets or Excel) and your phone calendar
• 15–30 minutes now; 5–10 minutes/month ongoing

Step-by-step (do this now)

1. Scan statements + inbox (5–10 min). Look for vendor names and repeated charges under $20.
2. Create your sheet (3 min). Columns: Service | Typical Amount | Frequency | Last Charge | Next Billing | Auto-renew (Y/N) | Cancel-by | Reminder Date | Notes.
3. Add top 8–12 items (5–10 min). Estimate next billing from last charge if unknown; flag auto-renew where likely.
4. Create calendar events (3 min per item). Title: “Action: [Service] renewal — keep/cancel.” Set primary alert 7 days before for annuals (3–5 for monthlies) and a follow-up 1 day before. Put cancel-by in the note.
5. Quick verify top 5 (5 min). Visit vendor billing pages to confirm cancel rules and adjust dates in sheet.
6. Optional AI check (3 min). Paste exported transactions into an AI prompt (below) to find anything you missed; verify manually.
7. Set monthly 5-minute recurring review. Reconcile new charges, remove canceled items, and mark any unknown vendor as “Investigate.”

Copy-paste AI prompt (use in ChatGPT or similar)

“I will paste lines from my bank statement or email receipts. Identify recurring subscriptions only. For each item output a CSV row: Service name, typical amount, billing frequency (monthly/annual/unknown), last charge date (if shown), estimated next billing date, likely auto-renew (Y/N), suggested cancel-by date (days before billing). Do not include one-off purchases.”

Metrics to track

• Number of subscriptions tracked
• Total monthly and annual spend on subscriptions
• Number of renewals avoided/cancelled per quarter
• Savings realised ($) over 3 months

Common mistakes & fixes

• Missing small charges — fix: sort by amount, search vendors <$10 explicitly.
• Wrong billing dates — fix: use last charge date and vendor account page to confirm.
• Sharing credentials with apps — fix: export CSVs or forward receipts to a dedicated inbox instead.

7-day action plan

1. Day 1: Download statements and search inbox for receipts.
2. Day 2: Create spreadsheet and add top 8–12 subscriptions.
3. Day 3: Add billing dates, auto-renew flags, and set calendar nudges.
4. Day 4: Run the AI prompt to cross-check (verify results manually).
5. Day 5: Reconcile bank statement for missed items and update sheet.
6. Day 6: Cancel 1–2 low-value subscriptions identified.
7. Day 7: Schedule the 5-minute monthly review on your calendar.

Your move.

Short answer: Yes — modern AI can create reminders that use context (calendar, device state, recent messages) and your location to deliver the right prompt at the right moment.

The problem: Time-only reminders fire when you asked for them, not when they’re useful. That creates noise and missed opportunities.

Why it matters: Contextual, location-aware reminders reduce friction for errands, meetings and follow-ups — turning reminders into action triggers instead of ignored notifications.

What works (quick lesson): Combine three things: 1) location or geofence, 2) context signal (calendar, email, time of day, recent call), and 3) a clear next action in the reminder text. That gives you relevance + clarity.

1. What you’ll need
   • A smartphone with location services
   • An assistant or automation app that supports geofences and context rules (built-in assistant, Shortcuts/Automations, or an automation app)
   • Permission to use your location and calendar for the app you choose
2. How to set it up — step by step
   1. Enable precise location and calendar access for the assistant/automation app.
   2. Create a geofence for the place you care about (grocery, office, post office) with a radius that avoids false triggers (start ~100–300m).
   3. Define the trigger conditions: arrive/leave location, calendar event present, or time-of-day window.
   4. Write the reminder as a clear action (who, what, where, one-step next action).
   5. Test with one trigger, watch behavior for 48 hours, then scale.

Copy-paste AI prompt to generate reminder templates and automations

Create three location-aware reminder templates for a smartphone automation system: 1) Grocery store arrival reminder listing 3 prioritized items; 2) Office arrival reminder to check unread emails from upper management; 3) Leaving home reminder to take keys and wallet. For each template provide: trigger (arrival/leave and time window), geofence radius, exact reminder text (clear one-step), and a short test procedure.

Metrics to track

• Reminder completion rate (% marked done)
• False-trigger rate (notifications that were irrelevant)
• Average follow-through time (time from reminder to completion)
• Weekly time saved estimate (minutes)

Common mistakes & fixes

• Too many reminders: consolidate and prioritize; keep max 3 per location.
• Vague prompts: use single-step, action-first phrasing (“Buy milk — 2L, brand”) instead of “Don’t forget groceries.”
• Privacy concerns: limit permissions, disable cloud sync if you prefer local-only processing.
• Location drift: tighten/loosen geofence radius after 24–72 hours of testing.

7-day action plan

1. Day 1: Enable permissions and create one test geofence + reminder.
2. Day 2: Run the test; log results (triggered? relevant?).
3. Day 3: Add two more context rules (calendar-aware, leaving-home).
4. Day 4: Tweak wording to single-action prompts.
5. Day 5: Measure completion & false-trigger rates.
6. Day 6: Reduce noise (drop low-value reminders).
7. Day 7: Review metrics and decide the next set of places or contexts to automate.

Your move.

— Aaron

5-minute win: Paste your best and worst email into an AI and run the “Calibration Pairs” prompt below. You’ll get a 10-point do/don’t list and three rewrites you can use today.

The problem: Adjectives like “warm” or “professional” are vague. That’s why AI outputs slip into generic marketing speak. You need measurable rules and contrast (what you are and what you are not).

Why it matters: A clear voice guide speeds approvals, reduces rewrites, and lifts response. Trackable guardrails (sentence length, reading level, banned phrases) make your brand sound consistent across email, web, and social—without you babysitting every line.

Lesson from the field: Calibration pairs beat adjectives. Show the AI one message you love and one you’d never send. Ask it to extract rules, not vibes. Then lock those rules as guardrails for every future prompt.

What you’ll need:

• 2 “good” samples and 2 “bad” samples (100–150 words each).
• 3 brand attributes (e.g., warm, clear, practical).
• 20–45 minutes and a place to save your one-page guide.

Step-by-step (from zero to a one-page Voice OS):

1. Calibrate with contrast: Use the Calibration Pairs prompt (below) with your good/bad samples. Expect a first-pass rule set and before/after rewrites.
2. Lock guardrails: Ask for concrete mechanics: average sentence length, reading level, punctuation rules (e.g., contractions allowed; 0 emojis; max one exclamation per 200 words), words to use/avoid, CTA formulas.
3. Generate the one-page guide: Request sections: 2-sentence voice summary; 6 dos/6 don’ts; mechanics; phrases to use/avoid; CTA patterns; examples for email, product blurb, social.
4. Create reusable templates: Ask for parameterized templates with fill-in brackets and word-count limits per channel.
5. Test with real audiences: A/B one email template this week. Keep everything else identical. Save baselines and deltas.
6. Iterate: Use the Auditor prompt (below) to score real copy against your guide. Fix drift and update the guide monthly.

Copy-paste prompt 1 — Calibration Pairs (start here):

“You are my brand voice calibrator. I’ll paste two short samples I approve and two I reject. Extract my voice as measurable rules. Output a one-page with: (1) 2-sentence voice summary, (2) 6 dos and 6 don’ts, (3) Mechanics: average sentence length target, reading grade target, contractions policy, punctuation and emoji policy, tone (3–5 traits), power verbs list (10), banned phrases list (10), (4) CTA patterns (3) each with an example, (5) 3 channel samples: email subject + 1-sentence body; product blurb 20–35 words; social post 15–25 words. Replace vague adjectives with specifics. If anything is generic, substitute a concrete example. After the guide, provide 3 before/after rewrites of my “bad” samples into my approved voice.”

Copy-paste prompt 2 — Guardrails (pin this at the top of future chats):

“Use as non-negotiable rules for all outputs in this thread. If my request conflicts with the guide, warn me and propose an aligned alternative. Keep sentences to the target average and reading grade. Enforce the banned phrases. End each draft with one clear, low-friction next step.”

Copy-paste prompt 3 — Voice Auditor (for quality control):

“Score the following draft against from 0–100. Show a bullet list of mismatches (specific lines and why). Then rewrite the draft to score 90+ while keeping facts intact. Return: (1) Score, (2) Issues, (3) Revised draft, (4) 3 alt subject lines or hooks.”

Insider guardrails that tighten output quality:

• Reading level target: Grade 6–8 unless you sell to specialists—then set Grade 9–10.
• Sentence length: average 12–16 words; avoid 35+ word sentences.
• Punctuation: contractions on; emojis off; max one exclamation per 200 words.
• Specificity rule: replace adjectives with numbers, examples, or named benefits.
• CTA formula: clear benefit + small proof + low-friction next step + time anchor.

Metrics to track (set a baseline before you change anything):

• Email: open rate, reply rate, unsubscribe rate, time-to-first-reply, average subject length.
• Web/product: CTA click rate, scroll depth, time on page, reading grade.
• Social: clicks per 1,000 impressions, comments per post, saves/shares.
• Workflow: time to draft, number of review cycles, percent of content using the guide.

Common mistakes and fast fixes:

• Vague outputs: Add banned phrases and power verbs. Ask: “Replace all adjectives with concrete examples or metrics.”
• Overly stiff or casual: Paste one on-brand sentence and say: “Match this rhythm and sentence length across the draft.”
• Too long: Enforce word-count ceilings by channel (subjects ≤7 words; blurbs 20–35 words; posts 15–25 words).
• Single-scenario overfitting: Include 3 channel samples and 3 CTA patterns in the guide.
• Drift over time: Run the Voice Auditor monthly and update the guide (v1.1, v1.2…).

One-week rollout plan:

1. Day 1: Run Calibration Pairs; publish (one page).
2. Day 2: Generate 3 templates each for email, product blurbs, and social. Save as snippets.
3. Day 3: A/B test one email template (50/50 split). Track opens, replies, unsubscribes.
4. Day 4: Rewrite one product page using the guide. Measure CTA click rate and reading grade.
5. Day 5: Post two social variants. Compare clicks per 1,000 impressions and comments.
6. Day 6: Use the Voice Auditor on three live assets. Fix drift. Update the banned phrases list.
7. Day 7: Share with your team. Pin the Guardrails prompt in your content workflow.

What to expect: A usable guide in under 45 minutes, 80–90% voice match on first pass, and faster reviews immediately. Refinement comes from real metrics, not more adjectives. Keep it to one page and iterate.

Your move.

That one-liner is solid. Short, local, and priced. Now let’s turn it into a 30-minute daily system that books reliable, in-person gigs—without endless scrolling.

5-minute quick win: Post two versions of your ad in different local groups today to see what pulls more replies.

• Version A: “Local to [Your Town]. Available Sat–Sun for odd jobs (handyman, moving help, event setup). $25/hr. Can bring tools. Interested?”
• Version B: “Local to [Your Town]. Same-day help within [Neighborhoods]. Odd jobs (handyman, moving help, event setup). From $25/hr. $5 off within 3 miles. Can bring tools. Interested?”

What to expect: Version B often lifts responses because it signals speed and proximity. Keep both live for 24 hours. Track replies and booked jobs.

The problem Most people send generic messages, chase low-pay leads, and don’t follow up—so they waste hours and miss the quick “yes.”

Why it matters A consistent, AI-assisted outreach loop increases response rate, cuts risk, and lifts effective hourly pay.

Field lesson Hyper-local detail + clear availability + price anchor beats long bios. The follow-up within 24 hours closes the gap.

What you’ll need

• Your town/ZIP and 2–3 nearby neighborhoods
• Two weekly time windows you can guarantee
• Base rate and minimum acceptable rate
• Notes app or simple spreadsheet to log leads
• Two or three references or past roles you can name

Step-by-step: the 30-minute daily pipeline

1. Productize your offer (5 min). List three fast, clear services and flat prices.
   • “Furniture assembly – $60 first item, $20 each additional.”
   • “Heavy lifting/moving help – $30/hr, 2-hr minimum.”
   • “Event setup/tear-down – $30/hr, weeknights and weekends.”
2. Create your message pack with AI (8 min). Generate: one-liner ad, 2–3 sentence intro, and three follow-up variants (same day, 24 hours, “last slot today”). Save them.
3. Search in bursts (7 min). Use 8–10 targeted phrases in local groups and classifieds. Prioritize posts from the last 24–48 hours, within 5 miles.
4. Send 5 personalized messages (7 min). Add one local detail (street, venue, or neighborhood) and a clear next step: “Can stop by today 5–7pm or tomorrow 9–11am. Which works?”
5. Follow-up and confirm (3 min). Same-day nudge to non-responders. For yeses, send a written confirmation: date, time, address, scope, rate, payment on completion, what you’ll bring.

Insider trick: Cluster your day by neighborhood. Offer a small “nearby” discount within 3 miles to fill gaps. You’ll raise effective hourly pay by cutting travel time.

Copy-paste AI prompts

• Local ad variants:“You are a concise assistant. I live in [Your Town, ZIP], serve [Neighborhood A/B], available [days/times]. My offers: [3 bullet services with prices]. Write 6 one-line ads with a local hook and a clear CTA. Include one ‘same-day’ version, one ‘within 3 miles $5 off’ version, and one for bad weather (yard/garage/indoor jobs). Keep each under 25 words.”
• Lead triage + reply builder:“Act as my gig lead screener. From this message: [paste their post or DM], extract date/time, location, tasks, pay, risks, and missing info. Draft a 2–3 sentence reply that confirms scope, proposes two time windows, and states payment on completion. If risky (no pay info or vague address), ask two clarifying questions before committing.”

Metrics to track (daily/weekly)

• Messages sent (target: 25/week)
• Response rate = replies/messages (aim: 30%+)
• Booked rate = meetings/replies (aim: 40%+)
• Show rate = jobs done/meetings (aim: 90%+ with confirmations)
• Effective hourly rate = (Pay – travel cost) ÷ total time door-to-door (aim: ≥ your minimum)

Common mistakes and quick fixes

• Vague scope → unpaid extra work. Fix: use a bullet list in your confirmation (tasks included, time estimate, what’s not included).
• Underpricing travel. Fix: price ladder (base rate + small travel fee beyond 5 miles) and neighborhood clustering.
• No follow-up. Fix: schedule a same-day nudge and a 24-hour check-in. Many yeses come on the second touch.
• Accepting low-quality leads. Fix: minimum rate and a 6-point vet checklist (pay method, address, parking/access, tools needed, time window, contact name/phone).

One-week action plan (crystal clear)

1. Day 1: Run both prompts. Finalize 3 productized offers, 6 ad variants, and your confirmation template.
2. Days 2–3: Post 2 ad variants/day in different groups. Send 5 personalized messages/day using fresh posts (last 48 hours). Track in a simple sheet.
3. Day 4: Follow up on all non-responses. Book 2 meetings. Test the “within 3 miles $5 off today” variant to fill gaps.
4. Day 5: Do the jobs. After completion, ask: “If you have a neighbor who needs help this week, I have one opening on [day/time]. Want an intro?”
5. Day 6: Repeat the 30-minute pipeline. Prioritize the highest-converting channel from your tracker.
6. Day 7: Review KPIs. Keep the best two ad variants, drop low-response channels, and adjust your minimum rate if effective hourly is low.

Expectation set: With 25 messages/week and two solid follow-ups, you should land 2–4 paid in-person gigs weekly in most areas. Quality improves fast as you refine offers, neighborhoods, and timing.

Your move.

Five-minute start: paste your latest transcript into an AI chat and use the prompt below to pull 3 hook quotes and one caption per quote. Post the best one today with a timestamp. That’s your first signal on what resonates.

Outcome: one episode → 3 clips, 2 social threads per clip, and a tight newsletter — created in a 60–90 minute block, scheduled for the week.

The snag: most repurposing dies in “where do I start?” and “what do I cut?” You need a simple AI brief that decides the angles for you so you’re editing, not guessing.

Why it matters: packaging beats polishing. Consistent, on-message micro-assets drive clicks and email signups without extra interviews. The compounding effect shows up in weekly CTR and subscriber growth.

Lesson learned: don’t ask AI for posts; ask it for a Repurposing Brief first. Then generate assets from that brief. It’s faster, more consistent, and easier to measure.

What you’ll need

• Episode audio/video and a transcript (with timestamps if possible).
• Basic tools: simple editor, audiogram/short-video creator, scheduler, and your newsletter platform.
• Your brand voice in one sentence, one primary CTA (listen/subscribe), and one secondary CTA (join newsletter/follow).

Step-by-step (one focused session)

1. Create the Repurposing Brief (10–15 min): paste the transcript into AI with Prompt 1. You’ll get: 3 clip picks with why they work, raw quotes, title-card copy, hooks, CTAs, and a newsletter outline.
2. Cut the clips (15–25 min): trim 3 segments (30–60s each). Light clean-up. Add a 3s title card and 3s outro with your CTA. Export captions if your tool supports it.
3. Generate the assets (15–20 min): feed each clip’s timestamps and brief back into AI with Prompt 2. You’ll get per-clip: 1 caption, 2 thread variants (LinkedIn and X), newsletter blurb, title-card text, and suggested on-screen captions.
4. Polish & verify (10 min): run Prompt 3 to check quotes match the transcript word-for-word. Adjust any captions for clarity and brand tone.
5. Schedule (10–15 min): stagger posts across 5–7 days. Use one trackable link format (e.g., add ?utm_source=[platform]&utm_content=[clip#]). Queue the newsletter for your usual send day.

Copy-paste AI Prompt 1 — Repurposing Brief

Prompt: “You are my content repurposing strategist. I’ll paste a podcast transcript. Produce a one-page Repurposing Brief with: 1) a one-sentence episode promise for a business audience over 40, 2) the top three 30–60s clip candidates with timestamps, the exact quote to feature, and why each will stop the scroll, 3) three hook lines per clip (plain English, 12 words max), 4) one primary CTA and one secondary CTA mapped to each clip, 5) title-card text (max 7 words) per clip, 6) a newsletter outline with a 2-sentence intro, 3 bullet takeaways, and one pull quote. Keep tone confident, clear, practical.”

Copy-paste AI Prompt 2 — Asset Generator

Prompt: “Using the Repurposing Brief and these timestamps [PASTE], create for each clip: A) one 12–18 word social caption, B) a 5-step LinkedIn thread (short sentences, numbered, include timestamp and CTA), C) a 6–8 tweet/X thread (first line is a bold hook; end with a clean CTA + timestamp), D) a 50–70 word newsletter blurb, E) title-card text (max 7 words), F) on-screen captions as 5–8 short lines. Quote exact wording from the transcript. Tone: direct, helpful, businesslike for 40+ audience. Output clearly labeled sections per clip.”

Copy-paste AI Prompt 3 — Quote QA

Prompt: “Check all quotes and on-screen captions against this transcript. Flag any words not present, suggest exact corrections with timestamps, and confirm that CTAs and claims are accurate and non-sensational.”

What to expect

• Week 1: 2–3 hours (setup templates). Week 2 onward: 60–90 minutes per episode.
• Not every clip will hit. The goal is one clear winner per episode to guide next week’s angles.

Metrics to track (weekly, by asset)

• Clips: 3s hold rate (>70%), 25% watch-through (>30%), click-through on post link (>1.5%).
• Threads: link CTR (>2–4%), saves/shares ratio (>0.5% of impressions), comments per 1,000 impressions (>3).
• Newsletter: opens (industry baseline), CTR to episode (>4–6%), reply rate (>0.3%).
• Conversion: new email subscribers and follows per episode; aim for steady week-over-week lift.

Mistakes & fixes

• Overlong clips: if watch-through <25%, cut to the payoff line and front-load the hook.
• No captions: add them; most people watch on mute.
• Vague CTAs: pick one action. Use the same CTA across assets for cleaner attribution.
• Misquotes: always run Quote QA.
• Posting all at once: schedule across the week to increase total reach.

One-week action plan

1. Today (15 min): run Prompt 1 on your latest transcript. Approve 3 clips.
2. Tomorrow (30–45 min): cut clips, add title cards/outros, export captions.
3. Same day (20 min): run Prompt 2 to generate captions, threads, newsletter blurb.
4. QA (10 min): run Prompt 3. Fix any quote mismatches.
5. Schedule (10–15 min): stagger posts over 5–7 days; queue the newsletter. Use trackable links.
6. Next Monday (10 min): review metrics, keep the best-performing angle, retire the worst, and adjust next episode’s clip selection accordingly.

Package first, polish lightly, measure weekly. That’s how one episode fuels your clips, threads, and newsletter without extra interviews.

Your move.

—Aaron