Win At Business And Life In An AI World

RESOURCES

  • Jabs Short insights and occassional long opinions.
  • Podcasts Jeff talks to successful entrepreneurs.
  • Guides Dive into topical guides for digital entrepreneurs.
  • Downloads Practical docs we use in our own content workflows.
  • Playbooks AI workflows that actually work.
  • Research Access original research on tools, trends, and tactics.
  • Forums Join the conversation and share insights with your peers.

MEMBERSHIP

HomeForumsPage 88

Jeff Bullas

Forum Replies Created

Viewing 15 posts – 1,306 through 1,320 (of 2,108 total)
  • Author
    Posts
  • in reply to: How does Facebook legacy contact work? #123454
    Jeff Bullas
    Keymaster

    Planning for your digital legacy is a sensible and proactive step to take.

    The Short Answer: A Legacy Contact can manage tributes on your memorialised profile, like pinning a post, but they cannot log in as you, read your private messages, or delete any of your past content.

    The feature is designed to preserve the public archive of your shared content formats—your photos, videos, and posts—while strictly protecting the privacy of your text-based conversations.

    After an account is memorialised, your Legacy Contact can perform a few specific actions. They can write a pinned text post that will stay at the top of the profile, often used for a final message or service details, and they can also update the main profile and cover image content. It’s important to understand that they never get access to the account itself; they cannot log in, see any of the private text content in your Messenger, or delete any of the past image or video content you have shared. When you set up the feature, you will also have the option to allow your Legacy Contact to download a copy of the photos and posts you’ve shared on your timeline.

    Cheers,

    Jeff

    Jeff Bullas
    Keymaster

    Dealing with buyer’s remorse is an unfortunate but common part of selling second-hand goods.

    Your Position: For an in-person, cash sale where the buyer inspected the item, you are under no obligation to provide a refund. Your best option is to politely state your position once and then cease communication.

    Your best defence in these situations is the quality of the original text and image content you created for the listing.

    You should send one final, polite message stating that the item was sold ‘as is’ and was accepted by them at the time of purchase, making the sale final. Do not get drawn into an argument; if they become abusive, report their messages and then block them. For future sales, be deliberate with your content formats. Your text content in the description should always include a phrase like ‘Sold as is, final sale’. Your image content should be comprehensive, with many clear photos from all angles, and even a short video demonstrating the item’s condition. This collection of text, image, and video content becomes your proof that you accurately represented the item.

    Cheers,

    Jeff

    Jeff Bullas
    Keymaster

    Dealing with a confusing ad rejection is a frustrating but very common rite of passage for advertisers.

    The First Steps: Your two options are to either make a small edit to the ad to trigger a fresh review by the system, or to directly request a manual review if you are certain you haven’t broken any rules.

    The automated review system is incredibly sensitive and often misinterprets an element within your ad’s image or text content, flagging it incorrectly.

    Before you do anything, carefully scrutinise every part of your ad’s content. Look at your main image or video; could anything be misinterpreted by an automated system? Then, read every single word of your text content, including the headline and description, as the system often flags single words or phrases that are commonly associated with policy violations, even if your use is completely innocent. If you find a potential issue, editing that specific part of the text or swapping the image and republishing will submit it for a new review. If you are confident your content is compliant, you should use the ‘Request Review’ option in Ads Manager to have it sent to a human for a final decision.

    Cheers,

    Jeff

    Jeff Bullas
    Keymaster

    Dealing with disruptive members is a common, and critical, challenge for any group admin.

    The Short Version: The best way to handle a disruptive member is to use a private, escalating process based on your group’s written rules, starting with a warning and leading to removal if the behaviour continues.

    This structured approach ensures you are being fair and transparent, which protects the health of your community by focusing on the disruptive text content, not the person.

    Your first step should always be to delete the offending comment and send the member a private message, clearly explaining which rule they broke and issuing a formal warning. Public arguments only fuel the fire and undermine your authority. If the behaviour continues, your next step should be to use Facebook’s admin tools to temporarily mute the member for a set period, like 24 hours or a week, informing them that this is their final warning. If they still cause problems after the mute expires, you should permanently remove and ban them from the group to protect the positive atmosphere for all other members.

    Cheers,

    Jeff

    Jeff Bullas
    Keymaster

    It’s a valid question, as the line between social and professional platforms has definitely blurred.

    The Short Version: Yes, for local service and retail roles like yours, it is a surprisingly effective tool for reaching a large pool of nearby candidates, and it’s free to post.

    The key to making it work is to treat your job listing as a compelling piece of content, using strong text and images to attract the right people.

    The biggest advantage is that you are advertising the role where potential local employees already spend their time. It’s free to post a job, and the simple application process means you will likely receive a high volume of candidates quickly. The main drawback to prepare for is applicant quality; because it’s so easy to apply, you may need to sift through more unsuitable candidates to find the good ones. My advice is to absolutely use it. Create your job listing with a clear text description and a good photo or video of your workplace, and be prepared to manage the applications as they come in.

    Cheers,

    Jeff

    in reply to: What does CTOR mean in email marketing? #123434
    Jeff Bullas
    Keymaster

    An excellent question. CTOR is a much sharper metric for measuring content effectiveness than the standard click-through rate.

    Short Answer: CTOR stands for Click-to-Open Rate. It measures the percentage of people who opened your email and then also clicked a link. It is calculated by dividing the number of unique clicks by the number of unique opens, not by the total number of emails sent.

    Think of it this way: your Open Rate measures the success of your subject line, but your CTOR measures the success of the actual content inside your email.

    The standard Click-Through Rate (CTR) is a measure of clicks against all the emails you delivered. This means it is influenced by your open rate; if nobody opens the email, nobody can click. CTOR is a more precise metric because it isolates the audience that actually saw your content. It answers the question: “Of the people who opened this email, what percentage found the content compelling enough to take the next step?”.

    This is an incredibly valuable diagnostic tool. For example, if you have a campaign with a very high open rate but a very low CTOR, it tells you that you wrote a brilliant, attention-grabbing subject line, but the text, images, or offer inside the email failed to deliver on that promise. Conversely, a low open rate with a high CTOR suggests that your subject line was weak, but the content of the email was actually very persuasive to the few people who saw it.

    To improve your CTOR, you need to focus on the body of your email. This means writing clearer and more persuasive text, using more compelling images, having a single and obvious call-to-action, and ensuring your design guides the reader toward the click. It is worth noting that Apple’s Mail Privacy Protection can artificially inflate your open numbers, which in turn can artificially deflate your CTOR. However, it remains a very useful metric for comparing the relative performance of your email content from one campaign to the next.

    Cheers,
    Jeff

    in reply to: Browse Abandonment vs. Cart Abandonment Emails #123430
    Jeff Bullas
    Keymaster

    An excellent question. Understanding this difference is key to implementing more sophisticated email automation.

    Short Answer: A cart abandonment email targets users who put items in their cart but did not buy, showing high purchase intent. A browse abandonment email targets identified users who only viewed a product but did not add it to their cart, showing a much lower, earlier stage of interest.

    The fundamental difference is the level of user intent, which means your email’s text, tone, and offer must be adjusted accordingly.

    A cart abandonment email is triggered by a very strong signal of intent. The user has found a product, selected it, and added it to their cart. They are very close to purchasing. Because of this high intent, your email text can be more direct and sales-focused. It should feature images of the items left in the cart and a clear call to action to “Complete Your Purchase”. You can use persuasive text that creates urgency or even introduces a small discount to close the deal.

    A browse abandonment email is triggered by a much weaker signal. The user has only viewed a product page. They are still in the consideration phase, not the decision phase. Because their intent is lower, your strategy must be much softer and more helpful to avoid feeling intrusive or creepy. The text of your email should not be a hard sell. Instead of “Buy Now”, the tone should be one of customer service.

    For the content of a browse abandonment email, the text could offer more information about the product they viewed, link to a helpful buyer’s guide, or show images of related best-selling items in the same category. The call-to-action text should also be lower pressure, such as “View Details” or “Explore the Collection” rather than “Buy Now”. It is a gentle nudge to continue their research, not a hard push to complete a sale.

    Cheers,
    Jeff

    Jeff Bullas
    Keymaster

    A very common and understandable goal for email marketers. It is, however, a complex issue.

    Short Answer: Getting emails to the Primary tab is not something you can directly control, as it is determined by Gmail’s algorithm and individual user behaviour. The best approach is not to try and ‘trick’ the algorithm, but to explicitly ask your subscribers to move your emails to Primary, which trains their inbox for future delivery.

    The promotions tab is not the same as the spam folder, and trying too hard to avoid it can sometimes do more harm than good.

    Gmail’s algorithm sorts emails by looking at thousands of signals, including sender information, email code, and the content itself. Any email sent from a major email marketing platform contains certain headers that identify it as bulk mail. Furthermore, a marketing email’s text and images, with words like “sale” or “discount” and a heavily formatted HTML layout, are strong signals that it is a promotion. Trying to game this by stripping out all your images and links just to look like a plain text email is a bad strategy; you sacrifice good design and clear calls-to-action, and the algorithm often sees through it anyway.

    The single most powerful signal you can send to Gmail is user action. Therefore, the most effective strategy is to ask your subscribers for help. You can run a simple campaign, especially for new subscribers, that asks them directly. The text can be as simple as, “To make sure you never miss our content, please drag this email from the Promotions tab to your Primary tab.” You can even include a simple animated GIF image showing them how to do it. When a user does this, they teach their personal inbox that your emails are important to them, which is far more powerful than anything you can change in your email’s design.

    It is also important to remember that people do check their Promotions tab when they are in the mood to engage with brands and offers. For many businesses, landing there is perfectly fine. The ultimate goal should be to send such valuable content that your readers will find it, regardless of which tab it lands in.

    Cheers,
    Jeff

    Jeff Bullas
    Keymaster

    A great question. In a world of online skepticism, visually communicating trust is more important than ever.

    Short Answer: You effectively use trust seals and security badges in an email by placing these small, recognizable images directly next to your primary call-to-action. Their purpose is to reduce a customer’s perceived risk at the exact moment you are asking them to make a decision.

    Think of these images not as decoration, but as a final, visual reassurance that it is safe for the reader to proceed.

    The psychology behind trust seals is that they act as a cognitive shortcut. When a potential customer is about to click your “Buy Now” button, they have a moment of hesitation where they assess the risk. A recognizable image from a trusted brand like Norton, McAfee, or even just a simple padlock icon, provides instant visual reassurance that the transaction is secure, helping to overcome that last-minute friction.

    Because of this, placement is the most important factor. These images have the most impact when they are placed directly adjacent to your main call-to-action button. Placing them in the footer of your email is far too low, and placing them at the top is too disconnected from the moment of decision. The ideal layout is to have your main call-to-action text, such as “Complete Your Secure Purchase,” immediately followed by a small row of trust seal images.

    There are a few types of seals to consider. First are security seals, like an SSL badge, that communicate technical security. Second are third-party seals, like a Better Business Bureau logo, that show your business has been vetted. Third are payment and guarantee seals, such as the logos of the credit cards you accept and a “100% Money-Back Guarantee” image, which reduce financial risk.

    A critical warning, however: you must only ever use official seals that you are actually entitled to display. Using a security brand’s logo without being their customer is deceptive and illegal. For a new business, starting with the logos of your payment processors and your own guarantee badge is a perfectly honest and effective way to begin building that visual trust.

    Cheers,
    Jeff

    Jeff Bullas
    Keymaster

    An excellent question that gets to the heart of sophisticated email design.

    Short Answer: Progressive disclosure in email is a design strategy that shows readers only the essential text and image content upfront, while keeping more detailed information hidden until it’s requested. This is typically achieved through ‘read more’ links or, in more advanced cases, interactive elements like accordions.

    The goal is to reduce cognitive load and make a long email feel scannable and manageable, enticing the reader to engage further rather than overwhelming them.

    The most common and reliable way to use progressive disclosure in an email is with a ‘teaser’ approach. For each section of content in your newsletter, you provide a strong headline, an engaging image, and a short, one-to-two sentence summary of the topic. This text provides just enough information to pique the reader’s interest. You then follow this with a clear call-to-action link with text like “Read the Full Story” or “Learn More”. This turns your email into a scannable table of contents that links out to your website, rather than a wall of text that nobody will read.

    A more advanced, but less supported, method is to build interactive elements directly into the email. Using clever code, it is possible to create things like an accordion or a toggle for an FAQ section. A user can click a question, and the text for the answer will expand right there in the email. This keeps the initial layout clean and allows the user to choose what they want to read.

    However, it is critical to understand that these interactive techniques are not supported by all email clients, most notably older versions of Outlook. If you use them, you must code a “fallback” so that the email still looks good and all the content is accessible in those clients. Because of this complexity, the simple and effective ‘teaser text’ with a ‘read more’ link remains the most practical strategy for most marketers.

    Cheers,
    Jeff

    Jeff Bullas
    Keymaster

    Quick 5-minute win: open your notes app and write three must-haves: daily protein target, one food you won’t eat, and max cooking time per meal. Use that to run an AI prompt below and get a usable 7-day plan in under 5 minutes.

    Nice tip — the three-must-haves trick is a brilliant way to keep inputs simple and actionable. Here’s a practical follow-up that turns that note into a real week of meals, shopping and small habits that save time.

    What you’ll need

    • Those three must-haves (age or rough activity level helps too).
    • A short list of dislikes/allergies and one weekly budget figure (optional).
    • A phone or laptop to run the AI and a notes app to capture the plan.

    Step-by-step — get a ready-to-shop week (30–60 minutes)

    1. Spend 5 minutes: write your three must-haves and one goal (maintain, lose 1 lb/week, or build/keep muscle).
    2. Copy-paste the prompt below into an AI tool and ask for a 7-day plan with two snacks, portion sizes, and a consolidated grocery list grouped by category. (One click, under 5 minutes.)
    3. Spend 10–20 minutes reviewing: swap any disliked recipes and ask the AI for simple swaps that keep the same cooking time and portions.
    4. Ask the AI to consolidate ingredient quantities into a single grocery list and round up amounts to avoid a midweek store run.
    5. Batch-cook once or twice (cook a grain, roast a protein, chop veg). Store in containers — enough for 3–4 days.

    Copy-paste AI prompt (use as-is):

    Create a 7-day meal plan for a 52-year-old female, lightly active, goal: maintain weight and improve heart health. Include breakfast, lunch, dinner and two snacks each day. Daily calories ~1800–2000, protein ~80g/day, fiber 25g/day, low saturated fat, no shellfish, includes 2 vegetarian dinners, and average cooking time 25–30 minutes per meal. Provide portion sizes, simple recipe notes, and a consolidated grocery list grouped by Produce, Dairy, Pantry, Meat/Fish, and Frozen with quantities for 7 days. Keep recipes repeatable (4–6 core recipes) to reduce prep time.

    Example — quick inputs:

    • Must-haves: 80g protein/day, no mushrooms, 30 min cooking max.
    • Result: AI returns a 7-day plan, portions, and a grocery list you can shop from immediately.

    Common mistakes & fixes

    • AI gives too many new recipes — Fix: ask for only 4–6 core recipes repeated with small variations.
    • Portions feel off — Fix: specify calories and protein per day in the prompt and ask for gram weights.
    • Grocery list is scattered — Fix: request a consolidated list grouped by aisle/category and rounded quantities.

    7-day action plan

    1. Day 1: Write your three must-haves and run the prompt (5–10 min).
    2. Day 2: Review, swap dislikes, and consolidate grocery list (20–30 min).
    3. Day 3: Shop and batch-cook staples (60–90 min).
    4. Day 4–7: Follow plan, note adherence and one feeling metric (energy or sleep).
    5. End of week: Adjust protein or recipes by 10–15% and run the prompt again.

    Closing reminder: Start small, repeat few recipes, and measure one thing (protein or how you feel). Small, consistent tweaks beat perfection. Try the prompt now — you’ll have a usable plan in minutes.

    Jeff Bullas
    Keymaster

    Short answer: Yes — with a good photo/scan and a few smart steps (including AI-assisted cleanup and vector tracing) you can turn hand-drawn lettering into clean, printable vector paths that scale perfectly.

    Why it works: vectors describe shapes mathematically, so once you trace your lettering into vectors you can print at any size without losing quality. AI helps by cleaning noise, increasing contrast and suggesting simplified paths — but expect some manual touch-up.

    What you’ll need

    • A high-resolution scan or photo (300–600 DPI or a smartphone photo with good lighting)
    • A raster editor or AI image cleaner (Photoshop, GIMP or an AI denoiser/background remover)
    • A vector editor: Adobe Illustrator or free Inkscape
    • Time for a short manual cleanup with the pen/vertex tools

    Step-by-step

    1. Capture — Scan flat at 300–600 DPI or photograph straight-on in even light. Avoid shadows and skew.
    2. Preprocess — Crop, straighten, increase contrast, convert to pure black & white (not grayscale). Remove specks.
    3. AI cleanup (optional but helpful) — Use an image-cleaning AI to remove texture, strengthen strokes and make the background pure white.
    4. Vectorize — In Illustrator use Image Trace: Mode = Black and White, adjust Threshold until strokes are solid; set Paths ~60–75%, Corners ~50–75%, Noise = 1–10 px, then Expand. In Inkscape use Trace Bitmap: Brightness cutoff/Edge detection, try Smoothing = 1–2 and Threshold until it matches.
    5. Refine paths — Use Simplify (reduce nodes), delete tiny artifacts, join endpoints, convert strokes to outlines (Object > Expand/Stroke to Path) if needed.
    6. Export for print — Save SVG/PDF/EPS and a high-res PNG for proofs. Check at large size to confirm smooth curves.

    Example quick workflow (Illustrator)

    1. Scan at 600 DPI → open PNG in Illustrator.
    2. Image Trace > Black and White > Threshold ~180 → Paths 70% Corners 60% Noise 2 → Trace → Expand.
    3. Simplify paths (Object > Path > Simplify), remove stray points, then File > Save As > SVG (for web) or PDF/EPS (for printers).

    Common mistakes & fixes

    • Low-res photo → rescan at higher DPI or retake photo with better lighting.
    • Too many nodes / jagged curves → use Simplify and manually adjust anchors.
    • Lost brush texture you wanted to keep → keep a high-res raster copy and combine raster texture with vector outlines.

    Copy-paste AI prompt (use with a text-to-image/assistant tool to clean your raster)

    “I have a high-resolution photo of hand-drawn lettering. Clean the image: remove background to pure white, increase contrast so strokes are solid black, remove specks and paper texture while preserving stroke edges and any intentional brush tails. Output a 3000–6000 px wide PNG at 300–600 DPI with transparency and a separate flattened PNG on white for tracing.”

    3-step action plan (today)

    1. Scan or photograph your lettering at high resolution.
    2. Run the image through the AI cleanup prompt above or adjust contrast manually.
    3. Open in Illustrator or Inkscape and follow the vectorize + refine steps.

    Expect to spend 10–30 minutes per piece for most cleanups, longer for detailed brush textures. Start simple, get one clean vector, and you’ll quickly build confidence. Vectors give you the freedom to scale, print and edit — the payoff is worth the small setup time.

    Jeff Bullas
    Keymaster

    Quick win (under 5 minutes): open a case study, copy the headline sentence, the top metric, and one short customer quote into a blank doc. That’s your one-pager skeleton — you can refine it after lunch.

    What you’ll need:

    • A full case study or call notes.
    • One clear outcome metric (revenue, time saved, % lift — pick the most tangible).
    • One concise customer quote to keep word-for-word.
    • A simple layout: headline, problem, solution, result, CTA.

    Step-by-step (do this now):

    1. Skim and extract (5 minutes): mark the customer’s pain, the action you took, and the measurable result.
    2. Pick the anchor metric (2 minutes): choose the single number that makes the most impact.
    3. Draft the one-pager (10–15 minutes): write a 7–10 word, benefit-first headline; one-sentence problem; one-sentence solution; then a results line that leads with the bolded metric and the verbatim quote beneath it.
    4. Create two variants (15 minutes): swap the headline focus or lead with a different metric for A/B testing.
    5. Polish for skimmers (5 minutes): bold the headline and metric, shorten lines, keep one column and one page.

    Example (plug-and-play):

    Headline: Retailer cuts stockouts 40% in 3 months

    Problem: Frequent stockouts were costing lost sales during peak season.

    Solution: Implemented predictive reorder rules and dashboard alerts.

    Result: 40% fewer stockouts in 3 months

    Quote (verbatim): “We finally stopped running out of our best-sellers — sales recovered immediately.”

    CTA: Book a 15-minute demo to see the dashboard in action.

    Common mistakes & fixes:

    • Too many claims — fix: use one clear metric and cite its source (customer or internal).
    • Changing the customer’s voice — fix: keep one quote word-for-word to preserve trust.
    • Weak CTA — fix: give a single, low-friction next step (15-minute call, one-click download).
    • AI hallucinations — fix: instruct the AI explicitly not to invent numbers or quotes.

    Copy-paste AI prompt (use exactly):

    Rewrite the following case study into a persuasive one-page layout that preserves the customer’s exact quote and primary metric. Output: a 7–10 word headline, one-sentence problem, one-sentence solution, a results line that bolds the metric, and a one-line CTA. Keep tone credible and specific. Do not invent numbers or change the quoted sentence. Case study: [paste full case study here]

    5-day action plan:

    1. Day 1: Pick one case study and extract headline, metric, quote.
    2. Day 2: Draft two one-pager variants (headline or metric swap).
    3. Day 3: Use the AI prompt above to tighten language; review and keep the quote.
    4. Day 4: Send variant A to a small list or attach to proposals; track opens and replies.
    5. Day 5: Send variant B, compare results, iterate on the winner.

    Remember: the fastest wins come from one number and one real human quote. Use AI as a sharpening tool — you stay in charge of the truth and the tone.

    Jeff Bullas
    Keymaster

    Nice point: I love your emphasis on getting the printer’s spec sheet and not trusting screen colours — that saves half the headaches right there.

    Quick win (under 5 minutes): paste this prompt into ChatGPT or your AI design tool to get 6 layout thumbnails you can pick from:

    AI prompt to paste:

    Generate 6 distinct thumbnail layout descriptions for a 3.5 x 2″ double-sided business card. Each thumbnail should include: placement of a centered vector logo on the front or left-aligned, suggested type hierarchy for name/title/contact, recommended font styles (one serif or sans name), a 2-colour palette using hex #1A73E8 and a complimentary neutral, and an idea for a tactile finish (matte, spot UV, embossed). Keep each description to 2–3 short sentences and label them 1–6.

    What you’ll need

    • Vector logo (.SVG/.EPS/.AI) or high-res PNG (300–600 DPI).
    • Exact text to print (name, title, phone, email, website).
    • Brand colours (Hex and Pantone if possible) and font names.
    • Printer specs: final size, bleed (3mm/1/8″ typical), paper stock, finish, file format.

    Step-by-step (do-first mindset)

    1. Run the quick win prompt to get 6 thumbnail ideas and pick 2 you like.
    2. Ask the AI for higher-res mockups of those 2 (front + back), telling it: CMYK, 300 DPI, include 3mm bleed, keep logo as vector.
    3. Open the chosen mockup in a vector editor (Illustrator or Affinity). Replace raster logos with your vector file, set document to CMYK, 300 DPI, add 3mm bleed and safe margins.
    4. Convert fonts to outlines or embed, add crop marks and export PDF/X-1a. Order a small proof run.

    Worked example

    • Goal: 3.5 x 2″ double-sided, centered logo front, contacts back in a clean sans.
    • Use the AI prompt to get thumbnails, pick one, request a high-res mockup and export layers. In your editor set CMYK, 3mm bleed, convert text to outlines, then export PDF/X-1a.

    Common mistakes & fixes

    • AI outputs RGB JPGs — fix: recreate layout in vector editor and convert to CMYK.
    • Low-res logo — fix: trace or ask a designer to vectorise; don’t print rasters under 300 DPI.
    • Text too close to edge — fix: move into safe area (6mm from trim recommended) and re-export.

    Simple 3-step action plan

    1. Try the quick AI prompt now and pick a favourite layout.
    2. Gather logo, exact text and printer spec sheet.
    3. Do the vector-cleanup + export PDF/X-1a and order one proof.

    Reminder: use AI for speed and ideas, not as the final file. A short manual pass in a vector editor plus a proof print will get you professional, print-ready stationery every time.

    Jeff Bullas
    Keymaster

    Quick win (try in 5 minutes): Add a short exemplar answer to your Google Form and paste 5 anonymized open responses into an AI prompt that returns scores in a single line — you’ll get consistent, copy-paste-ready grades fast.

    Nice point about anonymizing and using exemplars — that reduces bias and lifts AI consistency. Here’s a practical next step that keeps things beginner-friendly and adds a simple way to bring AI scores back into Google Sheets.

    What you’ll need

    • A Google account (Forms + Sheets).
    • An AI chat tool (ChatGPT or similar).
    • Optional: a mail-merge add-on for returning results by email.

    Step-by-step (easy to follow)

    1. Create the Google Form, enable quiz mode, and add mostly multiple-choice items for instant points.
    2. Add your open question(s). In Sheets create three columns: Accuracy (0–2), Clarity (0–2), Examples (0–1) and a SUM formula for a 0–5 score.
    3. Write a short exemplar answer (1–2 sentences) and paste it into the sheet so the AI can compare.
    4. Copy 5–20 anonymized open responses into the AI tool. Use the prompt below (copy-paste) and ask the AI to return scores in a single line per response so you can paste them back into Sheets.
    5. Paste the AI output into your rubric columns, combine with auto-graded totals from the Form, then spot-check ~10% for quality.
    6. If satisfied, use a mail-merge add-on or Sheets script to return scores and feedback to students.

    Copy-paste AI prompt (use as-is)

    Grade each anonymized student answer below against this rubric: Accuracy (0–2), Clarity (0–2), Examples (0–1). Exemplar answer: “{EXEMPLAR}”. For each input, return a single line in this exact format: ID | TOTAL(0–5) | Accuracy | Clarity | Examples | One-sentence reason | One improvement tip. Do not add extra text. Student answers: 1: “{ANS1}”; 2: “{ANS2}”; 3: “{ANS3}”.

    Worked example

    • Question: “Write a 2-line polite follow-up email.” Exemplar: “Hi Sam—just following up on my last email about next week’s meeting. Please let me know if next Tuesday at 10am works. Thanks!”
    • Batch 10 responses, run the prompt, paste back scores, add to the Form’s auto-grade for a full total.

    Common mistakes & fixes

    • Mistake: vague rubric — Fix: make rubrics concrete (key phrases, required elements).
    • Mistake: sending raw student names to AI — Fix: anonymize before pasting.
    • Mistake: trusting AI blindly — Fix: spot-check and tweak exemplar/rubric if scores drift.

    Action plan (today)

    1. Make a 5-question Form (15–30 mins).
    2. Collect 5 test responses and anonymize (5 mins).
    3. Run the AI prompt on the 5 open answers and paste results into Sheets (10–20 mins).

    Start small, validate with real responses, and keep human checks. Quick wins build confidence — then scale.

Viewing 15 posts – 1,306 through 1,320 (of 2,108 total)