Jeff Bullas

Quick win (5 minutes): turn one page of your SOP into a realistic, confidence-tagged quiz draft. Copy the prompt below, paste your topic, 3 outcomes, and 3–5 common mistakes, and you’ll have a ready-to-edit item bank before your coffee cools.

Copy-paste AI prompt (use as-is)

“You are an experienced instructional designer. Build a 6–question scenario quiz for [TOPIC] aimed at [ROLE]. Use these learning outcomes: [PASTE 3 OUTCOMES]. Base wrong options on these real mistakes: [PASTE 3–5 COMMON MISTAKES]. For each question, provide: (1) a short realistic scenario with constraints (time, policy, customer mood), (2) 3–4 answer options, (3) the correct answer, (4) a one-sentence rationale, (5) the mapped outcome, (6) a follow-up line: ‘Confidence: High/Medium/Low’. Keep language simple and job-focused. End with a 3-line summary showing which mistake each question targets.”

Why this works: you’re testing decisions under pressure, not memory. The confidence line helps you spot blind spots fast.

What you’ll need

• 3 measurable outcomes tied to a job task.
• 10–20 bullets (or a short SOP/transcript) that reflect how work is done today.
• 3–5 common mistakes your team actually makes.
• An AI chat tool and your quiz tool (Forms/LMS) or a simple spreadsheet.
• 3–5 people for a quick pilot.

Step-by-step (from zero to pilot)

1. List outcomes and mistakes: write 3 outcomes in plain English. Then list 3–5 frequent errors (from tickets, audits, call notes).
2. Run the prompt: paste your details and generate the 6 questions. Ask for simple wording and real-world details.
3. Light edit for reality: swap placeholders for your product names, policies, and phrases your best performers use.
4. Build the quiz:
   • If using a spreadsheet, create these headers: Question, Option A, Option B, Option C, Option D, Correct Option, Objective, Targeted Mistake, Rationale, Confidence (H/M/L).
   • Paste each item under those headers, then import to your quiz tool or copy-paste manually.
   • Add a short-answer item if one outcome needs explanation (e.g., “Draft the two-step response you’d say on the call”).
5. Pilot fast: run the quiz with 3–5 people. Time them. Record per-item accuracy and the chosen confidence level.
6. Tag and tune: mark any question with <60% accuracy or >60 seconds to answer. Fix wording or options. Keep distractors that reflect real mistakes.

Insider tricks (high-value)

• Mistake mining: paste anonymized notes and ask AI to extract the top 5 failure patterns with “why it happens.” Use those as distractors.
• Confidence gating: if someone answers wrong with “High” confidence, route them to a 2-minute micro-coach task (e.g., one role-play or a short script rewrite).
• Scenario shell: Situation → Constraint → Goal → Options (1 right, 2–3 common-mistake distractors). Keep options similar length; avoid giveaway words.

Worked mini-example

• Topic: Handling delayed orders on angry calls.
• Outcome: Within 3 minutes, de-escalate, confirm order details, and offer two compliant options.
• Scenario Q1: “Caller says, ‘This is ridiculous—I needed it yesterday!’ System shows backorder, ETA 3 days.” Options: A) Explain warehouse constraints. B) Apologize, reflect emotion, confirm order and preferred outcome. C) Offer full refund immediately. Correct: B. Rationale: defuse + clarify before proposing options. Targets mistake: premature solutioning or explaining operations.
• Short-answer: “Write the two compliant options you’d offer and your closing sentence.” Rubric: 1 point per compliant option; 1 point for clear, positive close with next step.

Quality bar (what good looks like)

• Every question maps to one outcome and one common mistake.
• Scenarios include a constraint (time, policy, mood).
• Options are plausible; the wrong ones mirror real errors.
• Language is simple, action-focused, and brand-consistent.

Measure what matters

• Learning: average score + outcome-level mastery (aim +15–20 points by second iteration).
• Calibration: confidence–accuracy gap (coach overconfident-wrong first).
• Behavior: role-play pass rate on first attempt (target 80%+).
• Job KPIs: the one metric this training should move (e.g., escalations -15% or handle time -8–12%).

Common mistakes & fixes

• Generic content — Fix: inject your policies, product names, and phrases from transcripts.
• Trivia questions — Fix: force a decision under a real constraint.
• Uneven options — Fix: make all options similar length and tone; remove obvious giveaways.
• No pilot — Fix: always test with 3–5 people; revise items <60% accuracy.
• One-and-done — Fix: iterate biweekly; retire items learners game or memorize.

3-day rollout plan

1. Day 1: Pick one task and write 3 outcomes. List 3–5 mistakes from real work.
2. Day 2: Run the prompt, edit for reality, build the 6-question quiz + 1 short-answer + confidence checks. Set pass criteria (e.g., 80% + role-play pass).
3. Day 3: Pilot with 3–5 people. Capture item accuracy, time-to-answer, and confidence. Revise weak items and roll out to the team.

Optional prompt (slides + facilitator guide)

“Create a 25-minute micro-training on [TOPIC] for [ROLE]. Include: 3 measurable outcomes, a 6-slide outline (2 bullets each) with 2–3 sentence speaker notes, 2 short role-play scenarios, and a 10-question skills quiz (8 scenario MCQ + 2 short-answer). For each question, provide the mapped outcome, correct answer, rationale, and 3 wrong options pulled from these mistakes: [MISTAKES]. Add a confidence check after each item and finish with a one-page facilitator guide (timings, materials, pass criteria). Keep it practical and plain English.”

Closing thought: AI gets you a sharp first draft fast. Your edge is the context you add and the tight feedback loop you run. Build, pilot, measure, tweak. Two short cycles beat one perfect plan.

Nice point — yes: AI is best as a quick copyeditor when you give it a single line to protect and clear constraints. That habit-friendly frame is exactly right. I’ll add a few practical tweaks so you get faster wins and avoid the common pitfalls.

What you’ll need

• A short paragraph (3–5 sentences) or a batch of similar paragraphs.
• A one-line “must keep” (the core fact or message the AI can’t change).
• A clear target: tone (friendly, formal), and a length goal (percent cut or max words).
• Your AI chat/editor tool and 5–10 focused minutes per paragraph.

Step-by-step—do this now

1. Read the paragraph and write the one-line core message (your guardrail).
2. Tell the AI the tone and exact length target (for example: “reduce by ~35%” or “under 25 words”).
3. Ask for 3 short rewrites that must keep the core message, use active voice, and simple words.
4. Pick the best option, then ask for two tiny variants: one more formal, one more casual.
5. Quick check: read the line aloud and confirm facts. If nuance is lost, ask the AI to restore a specific phrase.
6. Batch trick: process 3 similar paragraphs in one prompt using the same rules to save time.

Example

Original (bloated): “Given the current circumstances and after careful consideration of all possible avenues, we have determined that proceeding with the project at this time would not be advisable due to budgetary constraints.”

Tightened options (examples the AI should return):

• “After review, we will not proceed with the project due to budget limits.”
• “We won’t move forward with the project because of budget constraints.”
• “Budget limits mean we cannot proceed with the project now.”

Mistakes & fixes

• AI over-simplifies: Fix by asking “preserve nuance” or request to keep a specific phrase.
• Facts altered: Say “Do not change numbers, names, dates” or highlight them in your prompt.
• Too formal or too casual: Ask for a tone tweak and give a one-line example for the tone you want.

Copy-paste AI prompt (use this)

Edit the paragraph below to make it 30% shorter while keeping this core sentence unchanged: “[PASTE YOUR CORE SENTENCE HERE]”. Use active voice, plain words, and provide 3 distinct short rewrites. Also give one version that’s slightly more formal and one that’s slightly more casual.

Quick action plan (5–10 minutes/day)

1. Pick one paragraph each morning and run the prompt above.
2. Choose the best rewrite and copy it back into your doc.
3. After a week you’ll notice common patterns—next week, tighten your first drafts before using AI so you get even faster returns.

Closing reminder

Start small, protect the meaning, and use the batch trick. You’ll get clearer sentences fast, and the habit compounds — tighten one paragraph today, and your next draft will already be better.

Own your ideas. Let AI do the heavy lifting around structure, clarity, and polish — while you keep the voice and the thinking.

You want faster outlines and cleaner revisions without plagiarism risk. The safest path: use AI for planning, summarizing, and coaching. You do the rewriting, citations, and final fact checks. Think of this as a 20–40 minute loop per section that compounds quality every pass.

What you need on your desk

• Your topic in one sentence and 2–3 guiding questions.
• 3–6 credible sources with author, title, year, and page/paragraph notes.
• Your required citation style (APA/MLA/Chicago).
• A timer, a plain-text editor, and access to a plagiarism checker.

The safe, repeatable workflow

1. Outline fast (5 minutes). Ask AI for 3–5 sections with 2–3 bullets each. Treat this as a map, not final prose.
2. Digest sources (8–12 minutes). For each source you’ll use in a section, get a neutral summary you can verify. Capture claim, evidence, and a page/para reference.
3. Draft small, then rewrite (10–15 minutes). For one section, have AI produce a 1-sentence thesis and a 120–160 word draft using only your sources. Then you rewrite it: change order, swap examples, vary sentence length, and add your perspective.
4. Cite and verify (5–10 minutes). Insert citations to the specific sources you used. Check dates, names, numbers against originals. Run a plagiarism scan; if anything flags, rewrite those lines and recheck.

Insider tools that keep you original

• Compression–Expansion Paraphrase: First compress a source idea into 25–35 words. Then expand to ~130 words in your style using different structure and your examples. This breaks the pattern that triggers overlap.
• Source-to-Sentence Notes: For each paragraph you draft, list the specific source lines you relied on, plus how you transformed them. This makes citations easy and prevents accidental copying.
• “Cliché Sweep” pass: Ask AI to highlight generic phrases (e.g., “It is important to note,” “in today’s world”) and give fresher alternatives. Replace anything that feels like stock phrasing.
• Voice Brief: Before drafting, give AI 3–5 sample sentences you’ve written elsewhere. Ask it to keep structure simple, prefer active voice, and limit adjectives. You’ll still rewrite, but the draft will be closer to you.

Copy‑paste prompts you can use now

• Outline + first draft (safe):“You are my academic writing assistant. Topic: [insert topic]. My guiding questions: [Q1, Q2, Q3]. Sources I will use (author, year, title): [list]. Create a 4‑section outline with 2–3 bullet points per section. Then for Section 1, give a 1‑sentence thesis and a 130‑word draft paragraph that uses only the listed sources. Include parenthetical citations like (Author, Year). Do not invent facts or sources. Keep phrasing neutral and simple so I can rewrite it.”
• Source digest (fact‑first):“Summarize this source in 3 bullets in my own words. Bullet 1: main claim. Bullet 2: key evidence with page/paragraph. Bullet 3: how this supports or challenges my thesis. Do not quote unless I ask. Source: [author, year, title + page/para notes].”
• Cliché hunter (style pass):“Review this paragraph. Highlight generic or overused phrases. Suggest 3 fresher, precise alternatives for each. Keep meaning unchanged. Paragraph: [paste your draft].”
• Fact & citation checklist:“Extract every factual claim that would need a citation from the paragraph below. For each claim, suggest which of my sources is the best support. Note missing page numbers if needed. Paragraph: [paste]. Sources: [list].”

Mini example: compression–expansion in action

1. Compression (30 words): Studies find remote work can improve output when managers set clear goals and measure results; benefits depend on tools, expectations, and fair policies across teams.
2. Expansion (your rewrite, ~120 words): When teams are judged on outcomes, not hours, remote work often lifts productivity. People trade commuting for focus, and managers get clearer on goals. The gains aren’t automatic, though. Leaders need to spell out what “done” looks like, give access to the right tools, and apply policies consistently so remote employees aren’t second‑class. In other words: results first, clarity second, fairness always.

Common mistakes and fast fixes

• Letting AI invent citations. Fix: you provide full source details; use AI only to format.
• Paraphrasing too close to a source. Fix: compress to 1–2 lines, step away, then expand in new order with your examples.
• Drafting entire sections at once. Fix: work one paragraph at a time; it’s easier to keep your voice and verify facts.
• Skipping a read‑aloud test. Fix: read each paragraph out loud; if you could hear a thousand people write it that way, make it more specific.

Quality bar: what “good” looks like

• Every paragraph links to at least one named source or clearly labeled personal reasoning.
• Quotes are short and rare, always in quotation marks with citations.
• AI‑generated sentences are all rewritten by you for rhythm, order, and example choice.
• Plagiarism scan shows only small, generic overlaps (e.g., titles, common terms).

One‑hour action plan

1. Pick topic and list 3–4 sources (10 minutes).
2. Run the outline + first draft prompt for Section 1 (5 minutes).
3. Rewrite that paragraph using compression–expansion and a read‑aloud pass (20 minutes).
4. Run the cliché hunter and fix generic phrases (10 minutes).
5. Insert citations, verify facts against sources, and run a plagiarism check (15 minutes).

Your next move: Share your topic, one key source (author, year, page/para), and the required citation style. I’ll produce a safe outline and a strictly sourced, short draft paragraph for you to rewrite into your own voice.

Nice quick-win — blocking 30 minutes tomorrow is the kind of small, immediate win that builds momentum. I’d add a couple of practical tweaks so that 30 minutes becomes repeatable: place it in your highest-energy window and label it with the specific task you’ll finish.

What you’ll need (small kit)

• A calendar you actually use (paper or digital).
• A short daily checklist (notebook or app).
• A timer (phone or kitchen timer).
• An energy map (note morning/midday/evening energy for one week).
• Your top 3 goals for the next 90 days, each with one measurable sign of progress.

Step-by-step 90-day roadmap (do this first, then repeat)

1. Days 1–7: Set your 3 goals, create the energy map, and block one 30–60 minute “Top Task” each weekday in your best energy slot.
2. Days 8–21: Add a second focused block 2–3x/week. Start a 5–10 minute evening review to plan the next day.
3. Days 22–45: Build habit stacks (attach a new tiny habit to an existing morning/evening routine). Do a 20-minute weekly review to track the measurable signs.
4. Days 46–75: Every two weeks pick one small optimization: shorten a meeting, delegate one task, or move a block to a better time.
5. Days 76–90: Evaluate progress on each goal, keep what works, drop what doesn’t, and celebrate wins.

Worked example (realistic for busy adults over 40)

• Goals: finish chapter outline (work), walk 20 minutes 4x/wk (health), call kids Sunday evening (personal).
• Energy map: high energy mid-morning, low mid-afternoon, steady evening.
• Weekly plan: Mon–Fri 9:00–9:30 Top Task (writing), Tue/Thu 11:00–12:00 Deep Block (research), Daily 9pm 5-minute review.

Common mistakes & fixes

• Do not overbook blocks. Fix: protect 1–3 small blocks, not your whole day.
• Do not measure only effort. Fix: choose one clear outcome per goal (pages written, walks completed).
• Do not be rigid. Fix: treat the plan as living — tweak every two weeks.

Action checklist (do / don’t)

• Do: Block time in your calendar today for tomorrow’s Top Task.
• Do: Track one simple metric for each goal.
• Don’t: Skip the weekly 20-minute review.
• Don’t: Expect perfect progress — expect adjustments.

Copy-paste AI prompt (use this to get a tailored 90-day plan)

Help me build a 90-day productivity roadmap. My top 3 goals are: [goal 1 with measurable sign], [goal 2 with measurable sign], [goal 3 with measurable sign]. My weekly calendar availability: [days/times available]. My energy map for a typical day: [morning/high or low, midday, evening]. Constraints: [caregiving, travel, meetings]. Give me a week-by-week plan, two focused blocks per weekday where possible, one weekly review routine, and one small habit to add every two weeks.

Quick reminder: Start with one tiny protected block tomorrow morning. Small, consistent wins create the momentum that lasts — especially when life and energy matter more than ever.

Hook: You don’t need to rip out Tableau or Power BI to get AI — you can add smart, practical features fast and with minimal fuss.

Quick correction: It’s a common worry that AI means rebuilding your BI stack. Not true. Use extensions, APIs, or embedded models to layer AI on top of your existing dashboards.

Why this works: Small, targeted AI features give big user-value: natural language Q&A, automated insights, anomaly alerts, and simple forecasting. These are low-risk, high-impact additions that business users understand and adopt quickly.

What you’ll need:

• Admin access to your BI tool (Tableau or Power BI).
• Access to your data source or an exported sample dataset.
• API access to an LLM or ML service (or use built-in BI extensions if available).
• A small budget for API calls or a developer to wire integrations (low-code is often enough).

Step-by-step plan:

1. Choose your first use-case — pick one quick win (natural language search, anomaly detection, or narrative insights).
2. Prototype with a sample dataset — export a small table and test your AI model against it (no production risk).
3. Connect via one of three paths: built-in features (Ask Data/ Q&A), extensions/plugins, or a simple API call from a dashboard action.
4. Create a friendly UI touchpoint: a chat box for Q&A, a highlighted insight card, or email/SMS alerts for anomalies.
5. Test with a handful of users, collect feedback, iterate, then scale access and governance policies.

Example — Add a Natural Language Insight panel to Power BI:

1. Export the key data table to a secure location or leave it in your data model.
2. Use Power BI’s AI visuals or a simple webhook that sends a filtered data slice to an LLM endpoint.
3. Return a short narrative: top 3 trends, anomalies, and a suggested action — display as a card in the dashboard.
4. Validate and refine prompts until the language is business-friendly.

Common mistakes & fixes:

• Mistake: Trying to automate everything at once. Fix: Start with one feature and measure adoption.
• Mistake: Sending raw sensitive data to an API. Fix: Mask or aggregate before sending; use VPC/private endpoints where possible.
• Mistake: Overly technical UI. Fix: Keep language simple and actionable — one insight, one action.

30/60/90 day action plan:

1. Day 0–30: Prototype one feature and get feedback from 5 users.
2. Day 30–60: Harden security, refine prompts, add UI polish.
3. Day 60–90: Scale to more dashboards, track usage and business impact.

Copy-paste AI prompt (use with your LLM):

“You are a data analyst. Given the following table summary: [briefly describe columns and latest metrics]. Provide: 1) Top 3 insights in plain English, 2) Any anomalies to investigate, 3) One suggested action. Keep each item concise and business-focused.”

Closing reminder: Start small, iterate fast, and measure adoption. AI in BI is about better decisions — not fancy tech. Try one feature this week and see how much easier it makes conversations.

Best, Jeff

Good point: focusing on engagement—not just content—is the smart way to make training stick. AI can accelerate that, but you still control the learning experience.

Here’s a practical, do-first playbook to create engaging training materials and quizzes using AI. Short, actionable steps you can use today.

What you’ll need

• Clear learning goals for the session (3–5 outcomes).
• Basic source material or subject-matter notes.
• An AI text tool (chat assistant) and a slide or document editor.
• A quiz tool (Google Forms, LMS quiz, or paper) and a small pilot group.

Step-by-step process

1. Define objectives: Write 3 clear learning outcomes (what learners should be able to do).
2. Create an outline: Ask the AI to produce a short lesson plan with time per section and activities.
3. Generate content: Use the AI to draft speaker notes, slide bullets, and short examples or stories.
4. Build quizzes: Ask the AI to create 8–12 mixed-format questions (MCQ, scenario, true/false, short answer) that map to each objective.
5. Human edit: Trim language, add company-specific examples, and ensure accuracy.
6. Pilot: Test with 3–5 people, collect feedback, and refine questions and pace.
7. Deploy & iterate: Run the session, review quiz analytics, and improve content monthly.

Practical example

Topic: Handling Difficult Customer Calls

• Learning objectives: Calm the caller, identify the issue, offer two solutions, close positively.
• Slide outline from AI: 1) Why tone matters, 2) Listening techniques, 3) Script templates, 4) Role-play exercises.
• Sample quiz question (AI-generated): “You’re on a call and the customer is shouting. What’s the best first response? A) Raise your voice B) Apologize and ask a clarifying question C) Put them on hold.”

Copy-paste AI prompt (use as-is)

“You are an experienced instructional designer. Create a 45-minute training outline on ‘Handling Difficult Customer Calls’ with 4 learning objectives, 6 slide titles with 2 bullets each, a 10-minute role-play activity, and 10 quiz questions (mix of multiple-choice, scenario-based, and short answer) mapped to each objective. Keep language simple and add two real-world script examples.”

Common mistakes & fixes

• Too generic content — Fix: give the AI company context and examples.
• Slides too text-heavy — Fix: convert bullets into 3-point visuals and add scenarios.
• Quizzes don’t measure skills — Fix: include scenario-based or short-answer questions.
• Blind trust in AI — Fix: always review and pilot with real learners.

7-day action plan

1. Day 1: Choose one training topic and write 3 objectives.
2. Day 2: Run the AI prompt to get an outline and slides draft.
3. Day 3: Edit slides and add company examples.
4. Day 4: Generate and refine quiz questions.
5. Day 5: Pilot with a small group and collect feedback.
6. Day 6: Adjust based on feedback.
7. Day 7: Deliver the session and review quiz results.

Reminder: AI speeds up creation — but keep a human in the loop. Test, tweak, and focus on how learners use the knowledge. Start small, iterate fast, and you’ll see quick wins.

Spot on: a ranked shortlist beats hunting for a single “perfect” time. Let’s turn that into a 5‑minute, no‑friction workflow you can use today, plus a couple of insider tricks to keep things fair over months, not just one meeting.

Try this now (copy‑paste prompt)

“Help me schedule a 60‑minute meeting across time zones. Participants and preferred windows (local time): Dana — New York, 08:00–18:00; Lee — London, 08:00–18:00; Priya — Bangalore, 09:00–19:00; Marco — São Paulo, 09:00–18:00. Avoid starting before 08:00 or after 19:00 local for anyone. Generate 8 candidate start times in UTC within the next 10 days. For each candidate: convert to each person’s local time; score fairness using this rule: start at 100, subtract 10 points for each person outside their window, and subtract 1 point per 15 minutes outside (floor at 0). Flag any daylight‑saving or holiday risks you notice. Return the top 3 ranked options with one‑line reasons, and provide a ready‑to‑send poll message with the 3 options in local times for each attendee. Also suggest a simple rotation rule for recurring meetings so no region gets more than 25% of slots outside their window over 4 meetings.”

Why this works

• You define fairness once; the AI does the math every time.
• Everyone sees their local times up front, which kills the back‑and‑forth.
• A rotation rule avoids the same team always taking the hit.

What you’ll need

• Names and cities (or precise time zones) for each participant.
• Each person’s preferred working window (earliest/latest) or a 2‑hour “best block.”
• Meeting length and any hard constraints (no Fridays, must end by 17:00, etc.).

Step‑by‑step

1. List the people, cities, and windows in one note.
2. Paste the prompt above into your AI tool and run it.
3. Skim the top 3 ranked options, check any DST flags, and pick 2–3 to propose.
4. Send the AI’s poll text as a single message and ask for a 24‑hour reply.
5. Book the winning slot. Save the fairness score so you can rotate next time.

Worked example (what you should expect back)

• Three options like “13:00 UTC” with conversions (e.g., 08:00 New York, 13:00 London, 18:30 Bangalore, 10:00 São Paulo), each with a fairness score and a short note such as “Priya +30 minutes past window.”
• A clear nudge like “Rotate burden to APAC next meeting; keep within 60 minutes of window for all.”
• A poll you can paste into email or chat listing Option A/B/C with everyone’s local times.

Insider upgrades for recurring meetings

• Weight key roles: Add “Weight Dana (host) x1.5 in fairness scoring; if ties, prefer options that keep Dana within window.”
• Shoulder‑hours budget: Add “Track minutes outside window per person; ensure no attendee exceeds 90 minutes outside window across 4 meetings.”
• Holiday awareness: Add “Flag if any option lands on common public holidays in those cities and suggest an alternate within 48 hours.”

Premium prompt for an 8‑week rotation

“Design an 8‑week, biweekly 45‑minute meeting for these participants: New York (08:00–18:00), London (08:00–18:00), Bangalore (09:00–19:00), Sydney (08:00–18:00). Produce 6 candidate start times per meeting, convert to local times, apply this fairness rule: start 100, minus 10 per person outside window, minus 1 per 15 minutes outside (floor 0). Choose the best slot each week so (a) average fairness ≥75, (b) no region has more than 2 meetings outside their window, (c) rotate which region is closest to 09:00 local. Output a simple schedule for 8 weeks with UTC starts, local conversions, fairness scores, and a note on who’s bearing the inconvenience each time. Flag DST changes and adjust within ±60 minutes if needed.”

Common mistakes and quick fixes

• Mistake: Using region codes (e.g., “ET”) without cities. Fix: Give city names; the AI will handle DST correctly.
• Mistake: Asking for one “best” time. Fix: Always ask for 3–5 options and a rotation rule.
• Mistake: Ignoring small overruns. Fix: Cap meetings at the hour and ask the AI to avoid end‑of‑day starts that push past windows.
• Mistake: Sharing full calendars. Fix: Collect a 2‑hour best block from each person instead.

Action plan

1. Today: Collect cities and preferred windows from your group (2 minutes).
2. Today: Run the quick prompt and send the AI‑generated poll with 3 options (3 minutes).
3. This week: Book the slot, note the fairness score, and record who was outside their window.
4. Next meeting: Use the rotation rule to shift the burden and keep average fairness at 75 or higher.

One more high‑value tweak

• Ask the AI to include a “backup within 48 hours” for each option. If someone declines, you switch instantly without restarting the whole process.

Shortlist, fairness score, rotation. Do this twice and you’ll never go back to manual conversions.

Favicon-first is the quiet superpower. If your name and icon still read at 16 pixels, you’ve nailed Visual Fit. Pair that with a short, clear brief and AI will give you usable name+logo systems in one sitting.

• Do map tone to shapes and colors: soft = rounded forms, muted palette; bold = angular forms, high contrast.
• Do ask the AI to note letterform opportunities (e.g., hidden shapes in A, N, M, O, G).
• Do demand a monochrome variant and a 16px favicon mock for each logo direction.
• Do score names on Memorability, Pronounceability, Uniqueness, and Visual Fit (1–5 each).
• Don’t pick clever spellings that are hard to say or search.
• Don’t allow more than 2 primary colors plus 1 accent at concept stage.
• Don’t test only with designers; ask 5–10 real target users.

Insider trick: the Letter-to-Shape bridge

• Soft sounds (S, L, M) usually suit curves and low-contrast palettes.
• Hard sounds (K, T, G) often fit angles and stronger contrast.
• If the leading letter can become an icon (G, N, H, A, M), you’ll get a cleaner favicon and social avatar.

What you’ll need

• 50–100 word brand brief, 3 audience bullets, 3 tone words.
• Constraints: include/exclude words, color preferences, domain considerations.
• 45–90 minutes total and a simple 1–5 scoring grid.

Step-by-step (run this once, end with a shortlist)

1. Prep (15 min): Write the brief, audience bullets, tone words, and constraints.
2. Combined brainstorm (30–45 min): Use the prompt below to generate 15–20 names + 6 logo directions with palettes, monochrome, favicon, and 1–2 taglines each.
3. Score (15–20 min): Rate each name 1–5 on Memorability, Pronounceability, Uniqueness, Visual Fit. Keep the top 6.
4. Refine (30 min): For your top 3 names, request three logo forms each: wordmark, icon+wordmark, emblem. Ask for black/white and 16px/32px favicon previews.
5. Validate (1 day): Show the 3 name+logo combos to 5–10 target users. Ask one question: “Which feels most like a solution for you?” and collect one-line reasons.

Copy-paste AI prompt (replace bracketed items):

“Create 20 brand names and 6 matching logo concept directions for a [industry/service]. Brief: [50–100 words on product, benefit, differentiator]. Audience: [3 bullets]. Tone: [3 words]. Constraints: include/exclude words [list], preferred colors [list]. For each name: give a one-line rationale, pronunciation hint, and 1–5 uniqueness score. For each logo direction: specify layout (wordmark/emblem/icon), 2 color palettes, monochrome and favicon notes (16px and 32px), and 2 short taglines that pair with the name. Also point out any letterform/icon opportunities (e.g., hidden shapes in initials) and provide 3 short image-generator prompts for initial visuals.”

Worked example (so you see the level of output to expect)

Brief: “Home energy audit and smart upgrades for homeowners 40+. We cut bills and carbon with clear guidance and vetted installers. Trustworthy, practical, premium.”

• Shortlist names: GlowGuard (mem 4/5), WattNest (mem 4/5), HearthGrid (mem 3/5), FrugalWatt (mem 3/5), EverHaven Energy (mem 3/5), CosyCurrent (mem 3/5).
• Pick to explore: GlowGuard — G can form a shield icon; suggests protection and efficiency.

• Logo concept A (icon+wordmark): Rounded shield-shaped G. Palette 1: deep navy + soft amber. Palette 2: forest green + warm gray. Monochrome: solid mark, no thin gaps. Favicon: shield-G at 16px with simplified inner cut.
• Logo concept B (wordmark): Humanist sans with open counters; slight curve on the G crossbar to suggest flow. Palettes as above. Favicon: stylized double-G monogram.
• Logo concept C (emblem): House outline with a rising bar inside forming a G pathway. Keep strokes thick for small sizes.

• Taglines: “Lower bills, lighter footprint.” / “Comfort that pays you back.”
• Image-generator prompts (paste into your logo tool):
  • “Minimal emblem logo for ‘GlowGuard’: shield-shaped letter G, clean geometric, deep navy and soft amber, flat vector, high contrast, suitable for 16px favicon, no gradients.”
  • “Wordmark logo ‘GlowGuard’ in humanist sans, gentle rounded corners, premium yet friendly, monochrome black on white, spacing optimized for small sizes.”
  • “Icon logo: house outline forming a letter G path inside, bold strokes, forest green and warm gray, flat design, exportable as SVG.”

Mistakes to avoid (and quick fixes)

• Generic names — Require a uniqueness rationale and score; ask for 3 alternatives that push metaphor further.
• Logos that break at small sizes — Force a 16px/32px preview; remove thin lines and micro-detail.
• Color overload — Cap at 2 primaries + 1 accent; test monochrome first.
• Ambiguous spelling — Add a pronunciation hint and a “say it aloud” check.

Metrics to track

• Time to first usable name+logo: under 24 hours.
• Preference in a 5–10 person test: aim > 60% for the top combo.
• Memorability (1–5): target > 4 for the chosen name.
• Small-size clarity: both icon and wordmark readable at 16px/32px.

7-day quick plan

1. Day 1: Run the combined prompt; shortlist 6 names.
2. Day 2: Generate 3 logo forms for top 3 names; include monochrome + favicons.
3. Day 3: Make 3 mockups (social avatar, favicon, business card corner).
4. Day 4: Run a 5–10 person preference test; collect one-line reasons.
5. Day 5: Refine the winner; tighten spacing and color.
6. Day 6: Ship a simple landing or social post; track clicks and saves.
7. Day 7: Review metrics; decide to iterate or lock and build.

Refinement prompt (copy-paste after shortlisting):

“Using my top 3 names [list], produce three logo variations per name (wordmark, icon+wordmark, emblem). For each: supply black-on-white and white-on-black, a 16px/32px favicon mock note, spacing guidance, and whether the leading letter supports a monogram. Flag any legibility risks and propose one simplification. End with a one-paragraph brand usage note (avatar, signage, small print).”

Start with Visual Fit, force the favicon, and let AI do the heavy lifting. You’ll get a cohesive voice-and-face in days, not weeks.