<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Lena Mintz on Medium]]></title>
        <description><![CDATA[Stories by Lena Mintz on Medium]]></description>
        <link>https://medium.com/@lenamintz?source=rss-7848b23da1e6------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*NZEvfXqtxLx4RVUrLByxCA.png</url>
            <title>Stories by Lena Mintz on Medium</title>
            <link>https://medium.com/@lenamintz?source=rss-7848b23da1e6------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 18:05:23 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@lenamintz/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[I Set Up My First ChatGPT Ad Campaign. Here’s What Actually Happened]]></title>
            <link>https://aiadvances.org/i-set-up-my-first-chatgpt-ad-campaign-heres-what-actually-happened-49433906e72c?source=rss-7848b23da1e6------2</link>
            <guid isPermaLink="false">https://medium.com/p/49433906e72c</guid>
            <category><![CDATA[marketing]]></category>
            <category><![CDATA[ecommerce]]></category>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[advertising]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Lena Mintz]]></dc:creator>
            <pubDate>Fri, 02 Oct 2026 16:16:12 GMT</pubDate>
            <atom:updated>2026-10-05T17:48:58.728Z</atom:updated>
            <content:encoded><![CDATA[<p>A few days ago, I opened ChatGPT Ads Manager and tried to launch a real campaign — the kind of thing I suspect a lot of small business owners will be doing very soon. Not a demo, and not a hypothetical built for a screenshot. For the experiment, I used <a href="https://mr-mintz.com/">Mr. Mintz</a>, my small e-commerce brand selling printable crafts and activities for kids. Many of the digital products cost around $3, which makes the economics of paid acquisition unusually unforgiving.</p><p>By the end I had four live ad groups, a verified advertiser account, a $25 daily budget, and my first impressions. I also had quite a few questions.</p><figure><img alt="ChatGPT Ads Manager showing four live Mr. Mintz ads for Handprint, Halloween, Christmas, and Thanksgiving crafts. All four ads are marked Serving with Increase Bid recommendations; Christmas has 2 impressions and Thanksgiving has 8 impressions at this stage of the experiment." src="https://cdn-images-1.medium.com/max/1024/0*_agltOjfNIyuPolb.png" /><figcaption><strong>The campaign at publication time: all four ads serving, with the first 10 impressions coming in despite the $0.10 max CPC.</strong></figcaption></figure><p>The interesting part wasn’t where the buttons were — most of the interface is familiar enough if you’ve ever touched Google or Meta Ads. The interesting part was working out what it actually means to advertise inside a conversation.</p><p><strong>First surprise: there aren’t really “keywords”</strong></p><p>My instinct was to think in search-advertising terms. If I want to sell a printable Thanksgiving craft, I know the searches I care about: easy Thanksgiving crafts for kids, turkey handprint craft, Thanksgiving classroom activities.</p><p>ChatGPT Ads has something that looks similar, called context hints, but treating them as keywords would be a mistake. OpenAI describes them as a way to help its systems understand the conversations, needs and situations where an ad might be relevant — not exact-match keywords, not audience-targeting rules, and not a guarantee that your ad appears for a particular phrase.</p><p>That distinction changes how you build the campaign. Instead of asking what someone will type, I found myself asking what someone might be talking to ChatGPT about when my product would genuinely solve their problem. For the Thanksgiving ad group, that produced hints like:</p><ul><li>easy Thanksgiving crafts for kids</li><li>printable Thanksgiving activities for kids</li><li>turkey handprint crafts for kids</li><li>Thanksgiving classroom activities for teachers</li><li>low-prep Thanksgiving crafts for kindergarten and elementary classrooms</li><li>parents looking for easy Thanksgiving activities to do at home</li><li>teachers looking for printable Thanksgiving crafts for students</li><li>Thanksgiving activities for kids before the holiday break</li></ul><figure><img alt="ChatGPT Ads Manager settings for the Thanksgiving Crafts ad group, showing a manual maximum CPC bid of $0.10 and context hints for Thanksgiving crafts, printable activities, handprint crafts, preschool and elementary school activities, and classroom use." src="https://cdn-images-1.medium.com/max/1024/1*YqzWUwcmqzKqIKSFkyS9BA.png" /><figcaption><strong>ChatGPT Ads combines bidding with “context hints” — descriptions of relevant conversations that guide matching without acting as exact-match targeting rules.</strong></figcaption></figure><p>Those aren’t eight keywords bidding against eight searches. Together they describe an intent space. It sounds like a small semantic distinction. I don’t think it is.</p><p><strong>I split one small store into four intents</strong></p><p>Rather than build one broad “kids crafts” ad group, I used the structure that already existed in the store. Mr. Mintz is organized largely around seasonal and holiday collections, so I chose the three upcoming ones — Halloween, Thanksgiving and Christmas — and added Handprint Crafts as an evergreen category. Each ad group pointed to its existing collection page and got its own context hints, creative and tracking parameters.</p><p>That also follows OpenAI’s own guidance — an ad group should revolve around a single product category, theme or customer need, with meaningfully different use cases split out when they need different messaging or landing pages. For a large advertiser that’s elementary. For a small store with hundreds of inexpensive products it matters enormously: I don’t want to pay for someone vaguely interested in “things to do with kids.” I want a reasonable chance of showing up when the conversation has commercial intent close to something I actually sell.</p><p><strong>The recommended bid made no sense for my economics</strong></p><p>This was probably the most important decision in the whole setup. OpenAI currently recommends starting Clicks campaigns at a maximum CPC of roughly $3-$5. My products often cost around $3, so you can see the problem</p><p>If a $3 product converts at 3%, the revenue break-even CPC before fees and every other cost is about nine cents ($3 × 3%). Even at a spectacular 10% conversion rate it’s thirty cents. Paying several dollars for a click to sell a $3 digital craft isn’t an acquisition strategy — it’s a very efficient way of transferring money from me to an advertising platform.</p><p>So I ignored the recommended range, chose Manual max bid, and set my maximum CPC to $0.10. Ads Manager promptly labelled it “Low bid,” which is fair enough. OpenAI itself notes that its automatic Maximize Results strategy doesn’t guarantee any particular CPC, CPA or ROAS, and points advertisers to manual controls when they need a specific bid limit — which is exactly my situation. I’d rather find out that ten cents can’t win enough auctions and raise it deliberately than start at $3.50 and learn the same lesson after paying for it.</p><p><strong>The minimum budget was $25 a day</strong></p><p>For USD campaigns, OpenAI currently requires a minimum daily budget of $25, and that $25 is shared across all my ad groups — it isn’t $25 for Halloween plus another $25 for Thanksgiving and so on. So the initial setup came out as: Clicks objective, United States, $25 a day, manual max CPC of $0.10, four ad groups.</p><p>Technically, OpenAI treats that daily budget as a seven-day average rather than a hard daily spending cap.</p><p>At that ceiling, $25 is theoretically as many as 250 clicks, if the whole budget were spent at exactly ten cents each. That’s arithmetic, not a forecast — real delivery depends on auction eligibility and competitiveness, and a low bid may keep the campaign from spending anything close to its budget. Which is precisely what I want to learn.</p><p><strong>Building the ads was the easy part</strong></p><p>The ads themselves are compact. The Halloween version is just “Halloween Crafts for Kids” with “Printable crafts for home or the classroom,” linking straight to the Halloween collection rather than the store homepage. Same principle for Thanksgiving, Christmas and Handprint Crafts: specific intent, specific collection, simple creative. OpenAI’s own advice runs along the same lines — concise creative and a directly relevant, reachable landing page.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*393c_dxQyQVQBn9zCehU1w.png" /><figcaption><strong>The Handprint Crafts ad: a simple message, a category-specific landing page, and creative that shows the activity itself.</strong></figcaption></figure><p>The Handprint Crafts image was the one I thought about most. Instead of a polished finished product, I used a close-up process shot: a child’s open palm, a paintbrush applying white paint, the printable underneath. It explains the product before you’ve finished reading the ad.</p><p><strong>Then came verification</strong></p><p>Ads Manager supports two verification paths: individual advertisers can complete personal verification, while registered businesses and organizations can complete business verification. The important thing is to choose the advertiser type that matches how you actually operate. Once verification was completed, the campaign could move forward.</p><p>The account also needed billing set up, and I saw a $50 temporary authorization on the card during the process. That wasn’t my $25 budget being charged in advance; it was part of payment authorization. Once verification cleared, the status changed, and eventually all four ad groups showed the word I’d been waiting for — Serving — next to the one I’d been warned about, Low bid.</p><p><strong>And then: two impressions</strong></p><p>The first time I saw the campaign actually move, it had two impressions. Not two thousand — two. By the time I was finishing this article the next day, those two impressions had become ten — and one of them had turned into a click. Total spend: $0.10. CPC: exactly $0.10, my maximum bid.</p><p>This is obviously far too little data to tell me anything about performance. But it does answer the first question I had: a $0.10 bid can actually get delivery.</p><p>My Thanksgiving group registered them first, and oddly, those two impressions were more useful to me than a dashboard full of projections. They proved that a campaign with my deliberately tiny ten-cent maximum wasn’t technically shut out of the auction — the system could serve it. That doesn’t tell me the campaign will work, or even that $0.10 will generate real volume. It tells me the experiment has actually started.</p><p>OpenAI says impressions, clicks and CTR refresh roughly every 15 minutes in Ads Manager, while spend can lag by seven or eight hours, so I’m not about to optimize anything off the first couple of hours. For now, the low bid stays.</p><p><strong>Three days later</strong></p><p>By October 4, the campaign had reached 33 impressions and four clicks across two ad groups, with $0.39 spent. Thanksgiving Crafts accounted for 21 impressions and two clicks, while Halloween Crafts produced 10 impressions and two clicks. Christmas had two impressions and no clicks, and Handprint Crafts had not delivered yet. There were still zero conversions.</p><figure><img alt="ChatGPT Ads Manager showing October 4 results for the Mr. Mintz campaign: Halloween Crafts received 10 impressions and 2 clicks with $0.19 spent and a $0.09 average CPC; Thanksgiving Crafts received 21 impressions and 2 clicks with $0.20 spent and a $0.10 average CPC. Christmas Crafts received 2 impressions and no clicks, while Handprint Crafts had not yet received impressions." src="https://cdn-images-1.medium.com/max/1024/1*TNcuSIAPZh9ghwxzaNgGjw.png" /><figcaption>Three days in: the $0.10 max CPC is still getting delivery, with four clicks across two ad groups and $0.39 spent.</figcaption></figure><p>Across the campaign, those four clicks cost $0.39 — just under $0.10 per click on average. The overall CTR was roughly 12.1%.</p><p>Those numbers look interesting. They are also almost meaningless at this sample size. Thirty-three impressions are nowhere near enough to conclude that I have found an unusually high-performing campaign, and four clicks are certainly not enough to infer the auction floor.</p><p>But there is one narrower conclusion I’m comfortable making: <strong>the first $0.10 click was not a one-off.</strong> A $0.10 maximum CPC has now produced delivery and clicks across two different ad groups. So for now, I’m leaving the bid exactly where it is.</p><p><strong>This is only Part 1</strong></p><p>I then went a step further and connected Mr. Mintz to ChatGPT Ads through Shopify. The integration synced 924 products, added the OpenAI Pixel and began sending real events from the store back to Ads Manager.</p><p>That’s Part 2 — because getting an ad to say <strong>Serving</strong> is one thing. Finding out whether any of those clicks actually turn into customers is considerably more useful.</p><p><strong>A note on how I wrote this</strong></p><p>This is my own experiment, with my own store, advertising budget and data. I used ChatGPT along the way to help me understand the new Ads Manager interface and work through the campaign setup. I also used it to help structure and edit this article in English. The store is mine, the decisions and money are mine, and the screenshots, results and observations come from my own experience.</p><p>I’m sharing it because I suspect I’m not the only small seller doing this math. When you sell low-priced digital products, a recommended cost per click can quickly approach — or exceed — the price of the product itself. If you’re trying to figure out whether a new advertising platform can make economic sense at that scale, I hope documenting the experiment as it actually unfolds is useful.</p><p><em>Tested in October 2026 with the OpenAI Ads Manager beta and a live Shopify store. Ads Manager is explicitly a beta product, so screens, features and workflows may change.</em></p><p><em>Originally published at </em><a href="https://sherunsai.substack.com/p/i-set-up-my-first-chatgpt-ad-campaign"><em>https://sherunsai.substack.com</em></a><em> on October 2, 2026.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=49433906e72c" width="1" height="1" alt=""><hr><p><a href="https://aiadvances.org/i-set-up-my-first-chatgpt-ad-campaign-heres-what-actually-happened-49433906e72c">I Set Up My First ChatGPT Ad Campaign. Here’s What Actually Happened</a> was originally published in <a href="https://aiadvances.org/">AI Advances</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[From Handprint Crafts to a Parenting Book: The Five-Year Evolution of Mr. Mintz]]></title>
            <link>https://medium.com/@lenamintz/from-handprint-crafts-to-a-parenting-book-the-five-year-evolution-of-mr-mintz-f71fcff19072?source=rss-7848b23da1e6------2</link>
            <guid isPermaLink="false">https://medium.com/p/f71fcff19072</guid>
            <category><![CDATA[entrepreneurship]]></category>
            <category><![CDATA[product-development]]></category>
            <category><![CDATA[parenting]]></category>
            <category><![CDATA[writing]]></category>
            <category><![CDATA[small-business-owner]]></category>
            <dc:creator><![CDATA[Lena Mintz]]></dc:creator>
            <pubDate>Fri, 18 Sep 2026 15:06:59 GMT</pubDate>
            <atom:updated>2026-09-18T15:06:59.270Z</atom:updated>
            <content:encoded><![CDATA[<h3>What started as a way to keep two children busy at home became a creative business — and eventually changed the way I think about building products.</h3><p>Five years ago, I wasn’t trying to build a children’s craft brand. My kids were little, we were spending a lot of time at home, and I was trying to solve a very ordinary parenting problem: I needed things we could actually do together. Not elaborate Pinterest projects requiring twelve materials and an adult with unlimited patience, but simple activities I could print, put on the table, and start.</p><p>So I began making them — handprint crafts, paper projects, little keepsakes, things designed around the reality of having actual children in the room. Some worked, some didn’t, and the ones that worked particularly well became templates I could share with other parents. That was the beginning of Mr. Mintz.</p><h3>We built what we needed first</h3><p>I came to this from a very different professional world. Before becoming an entrepreneur, I worked in corporate PR and communications, so I understood audiences, positioning and marketing. But none of that was really the starting point for Mr. Mintz. The starting point was much simpler: <strong>make what we need, and if it works, share it.</strong></p><p>Over time, that family project became <a href="https://mr-mintz.com/"><strong>Mr. Mintz</strong></a>, a real business whose printables now reach parents and teachers far beyond our own kitchen table. I learned something that seems obvious now but wasn’t obvious to me then: living inside the problem you are trying to solve changes the way you build a product.</p><p>You notice very quickly when an activity supposedly designed for children actually requires a parent to do most of the work. You learn that five extra minutes of preparation can be enough for a busy parent to abandon an otherwise lovely idea. You discover that instructions that look perfectly clear on a screen become much less clear when there is paint on the table and an impatient child waiting beside you.</p><p>We weren’t designing for an imaginary family. We were the family, and that became an unexpectedly useful product-development method.</p><h3>Then our customers grew up</h3><p>There is, of course, an obvious problem with building things for young children: children refuse to remain young.</p><p>Ours grew. The handprints got bigger, the activities that interested them changed, and eventually the problems we were trying to solve at home changed too. Instead of wondering what we could make together that afternoon, my husband and I found ourselves dealing with a completely different set of questions.</p><p>How much should we still manage for them, and how much should they manage themselves? Why could an ordinary request suddenly turn into an argument? When should we intervene and when should we back off? How do you give a child more independence without simply removing the structure they still need?</p><p>Most of all, we began noticing that methods that had worked perfectly well with younger children were becoming less effective as they grew older. It wasn’t necessarily that the children had become more difficult or that we had suddenly become worse parents. The relationship itself was changing, and we needed to change with it.</p><p>That realization eventually became <a href="https://link.amazon/B04YPwYnV"><em>On the Glide: Raising Kids Who Stay Close as You Step Back</em></a>, the book I wrote with my husband.</p><figure><img alt="Author Lena Mintz holding a copy of her parenting book, On the Glide: Raising Kids Who Stay Close as You Step Back." src="https://cdn-images-1.medium.com/max/1024/1*tYODzee9YgnoG1NFmxfdkg.png" /><figcaption><a href="https://link.amazon/B03b28BgR"><em>On the Glide: Raising Kids Who Stay Close as You Step Back</em></a></figcaption></figure><h3>A very different product built in a very familiar way</h3><p>On the surface, a parenting book about tweens has very little in common with a printable handprint craft. But while we were working on <em>On the Glide</em>, I kept recognizing the same process that had created Mr. Mintz.</p><p>We started with problems inside our own family. We paid attention to where things repeatedly broke down. We tried different approaches, kept what actually helped, discarded what didn’t, and gradually built a system around the patterns we were seeing.</p><p>One of the central ideas in the book is that as children move toward the tween and early teen years, the parent’s role has to change as well. You cannot manage a ten- or twelve-year-old exactly as you managed a five-year-old and expect the relationship to remain unchanged. More of the steering wheel has to move into the child’s hands, while the parent remains present, keeps the boundaries that still matter and provides a safe place to return when things go wrong.</p><p>That transition sounds straightforward when compressed into a paragraph. In a real family, of course, it is anything but straightforward.</p><p>Writing the book forced us to examine not only what we were doing with our children, but why. We started thinking much more deliberately about responsibility, conflict, trust, routines and what happens after mistakes. One question became particularly important to us: when something goes wrong, does the child still come back and tell us the truth?</p><p>That became one of the ideas at the heart of <em>On the Glide</em>. Connection, at this age, cannot simply mean the absence of conflict. Sometimes a better measure is whether the relationship can survive the conflict and allow the child to return.</p><h3>For a long time, I thought these were two separate stories</h3><p>I used to think of Mr. Mintz and my books as different parts of my professional life. One was a creative business built around children’s activities. The other was writing about work, family and, eventually, parenting.</p><p>Now I’m not so sure.</p><p>The same parents who once search for a handprint craft for a four-year-old eventually have an eight-year-old asking for more independence, then a ten-year-old with a device, then a twelve-year-old who wants more privacy. Families don’t remain inside neat market categories. They move.</p><p>Our own family moved too, and the things we were making moved with it.</p><p>That has started to influence the way I think about Mr. Mintz itself. For years, much of our work naturally centered on younger children because those were the children living in our house. Now we are increasingly interested in what useful creative materials might look like for older kids as well.</p><p>Not because somebody presented me with a particularly attractive demographic chart for ages 7–14. Our children simply got there first.</p><h3>Maybe the product roadmap was in front of me all along</h3><p>Business advice often begins with the market: identify demand, find an underserved niche, build something, test it and iterate. There is nothing wrong with that approach, and I use plenty of data in my business.</p><p>But looking back over the last five years, I can see another method running alongside it: <strong>make what we need, test it in real life, then share it.</strong></p><p>Being your own customer certainly has limitations. Your family is not a representative sample, and personal experience should never be confused with universal truth. But proximity to a problem has its own value. You see details that are easy to miss from a spreadsheet, because you encounter them at breakfast, after school, during an argument or at a kitchen table covered in paint.</p><p>Mr. Mintz began because our little kids needed things to make.</p><p><em>On the Glide</em> began because those kids didn’t stay little.</p><p>I don’t know yet what their next stage will make us build. After five years, though, I’ve stopped assuming that the answer will have nothing to do with my work.</p><p><strong>If this stage of parenting sounds familiar, you can learn more about </strong><a href="https://link.amazon/B0gn8VDkC"><strong><em>On the Glide: Raising Kids Who Stay Close as You Step Back</em></strong></a><strong>.</strong></p><p>I also recently had a long conversation with Wrote a Book about the book, Mr. Mintz, writing with my husband, and how these different pieces of my work ended up being much more connected than I originally realized.</p><p><strong>Read the full Author Spotlight </strong>→ <a href="https://wroteabook.org/author-spotlight-lena-mintz-about-on-the-glide/"><strong>Wrote a Book</strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f71fcff19072" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[I Left Corporate in a Pandemic. AI Just Made Me Rebuild the Whole Thing Again.]]></title>
            <link>https://medium.com/@lenamintz/i-left-corporate-in-a-pandemic-ai-just-made-me-rebuild-the-whole-thing-again-13c9cbc0e591?source=rss-7848b23da1e6------2</link>
            <guid isPermaLink="false">https://medium.com/p/13c9cbc0e591</guid>
            <category><![CDATA[entrepreneurship]]></category>
            <category><![CDATA[small-business-owner]]></category>
            <category><![CDATA[women-in-business]]></category>
            <category><![CDATA[midlife-women]]></category>
            <category><![CDATA[digital-entrepreneurship]]></category>
            <dc:creator><![CDATA[Lena Mintz]]></dc:creator>
            <pubDate>Wed, 17 Jun 2026 17:25:26 GMT</pubDate>
            <atom:updated>2026-06-17T17:25:26.044Z</atom:updated>
            <content:encoded><![CDATA[<p><em>At 50, I’m learning that “stable” was never the point.</em></p><figure><img alt="A sheet of paper divided into two handwritten columns, lit by warm lamplight on a dark table at night, with a pen and coffee cup nearby. Overlaid text reads “The night I chose the scarier column.” SheRunsAI logo in the corner." src="https://cdn-images-1.medium.com/max/1024/1*RkWKbnXz9gYVbJoX23BP6w.png" /></figure><p>In 2020, the night before I had to decide whether to go back to my corporate job, I sat at the kitchen table after everyone was asleep and drew two columns on a sheet of paper. Reasons to go back. Reasons not to.</p><p>I wasn’t thinking about purpose or reinvention. My maternity leave was ending, I had two small kids, and there was a pandemic outside. I was thinking about rent and kindergarten fees. The sensible choice was obvious: a stable salary at a large IT company. Walking away felt almost irresponsible.</p><p>But by the time both lists were done, one thing was clear. Going back meant choosing safety over a life that actually felt possible. I chose the other column. That quiet decision, made after everyone else had gone to sleep, turned into an Etsy shop, then a website, then the whole business I run now: <a href="https://mr-mintz.com/">Mr. Mintz</a>, digital craft templates for kids.</p><p>Six years later, I’m 50, and the ground is moving again.</p><p>This time it isn’t a pandemic. It’s AI, search algorithms, the way people find and buy things online, all shifting at once. And here’s the part that doesn’t fit neatly on a motivational poster: the same tools that let me rebuild a one-woman business are the ones that keep making last quarter’s playbook obsolete.</p><p>AI put a marketing team, a designer, and a research department on my laptop. Things that used to belong to companies with budgets now run on a subscription I can cancel anytime. That’s thrilling. It also means the thing I figured out in spring might not hold by autumn. I find it exciting and slightly destabilizing, usually in the same afternoon.</p><p>What’s changed at 50 isn’t the pace of the work. It’s me.</p><p>For years I treated exhaustion as proof I was working hard enough. I built things late at night because that was the only quiet, and I wore the fatigue like a credential. I’ve stopped. Not because I turned into a 5 a.m. wellness person (I didn’t), but because I finally noticed the math. Depleted, I work slowly and make mistakes I fix the next day. Rested, I do more in less time and do it better. The badge was costing me the thing it was supposed to prove.</p><p>So I’m trying to work like a long-distance runner instead of a sprinter who’s already out of breath. Serious, consistent, but not built on fumes.</p><p>There’s a quieter reason, too. My kids are 8 and 11. The window where they actually want me around is shorter than it feels. And the promise of AI, all that saved time, comes with a trap: it’s very easy to pour the hours you just freed up straight back into more work. I’m trying not to. It’s basically the argument I had with myself for a whole book, <a href="https://amzn.to/4aBQqgS"><em>On the Glide</em></a>, which came out in May.</p><p>I don’t have a tidy lesson here. Just a pattern I’ve made peace with: the structure you build is never finished. Every few years, life or business or technology asks you to reassemble it again, more honestly than before. At 50, that doesn’t scare me the way I expected. It mostly makes me want to sit down with a fresh sheet of paper.</p><p>I talked about all of this with Kiran Singh for her <em>85 Weeks to Fifty</em> series. <a href="https://kiransinghuk.substack.com/p/85weekstofifty-with-lena-mintz">The full interview is here.</a> And if you want the working version of how I actually use these tools to run a real business, I write about it at <a href="https://sherunsai.substack.com/">She Runs AI</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=13c9cbc0e591" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Three Ways Microsoft Copilot Found My Store — Without Me Doing Anything]]></title>
            <link>https://medium.com/@lenamintz/three-ways-microsoft-copilot-found-my-store-without-me-doing-anything-0f49d486e41a?source=rss-7848b23da1e6------2</link>
            <guid isPermaLink="false">https://medium.com/p/0f49d486e41a</guid>
            <category><![CDATA[ecommerce]]></category>
            <category><![CDATA[small-business]]></category>
            <category><![CDATA[shopify]]></category>
            <category><![CDATA[microsoft-copilot]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Lena Mintz]]></dc:creator>
            <pubDate>Wed, 22 Apr 2026 10:20:52 GMT</pubDate>
            <atom:updated>2026-04-22T10:20:52.496Z</atom:updated>
            <content:encoded><![CDATA[<p>Three Ways Microsoft Copilot Found My Store — Without Me Doing Anything</p><figure><img alt="Warm editorial illustration of a woman working at her laptop from home while an AI chat window shows three ways her store was discovered: a printable handprint product card, a comparison table, and a blog reference snippet, with small marketplace icons and a She Runs AI logo." src="https://cdn-images-1.medium.com/max/1024/0*H83ljwhy4hId41Nb.png" /></figure><p>On April 2, an email from Shopify landed in my inbox. The subject line was something about “Agentic Storefronts now supporting Microsoft Copilot.” I almost archived it. I get a lot of platform announcements, and most of them are either premature or irrelevant for someone running a small one-person business from a kitchen table.</p><figure><img alt="Screenshot from Shopify showing the “Agentic storefronts” section with the message “Your active products are automatically discoverable,” listing ChatGPT and Microsoft Copilot as connected AI discovery channels, with a She Runs AI logo in the corner." src="https://cdn-images-1.medium.com/max/1024/1*jr8L09JH3qoWVQ_8mgD9bw.jpeg" /></figure><p>But this one stayed open in a tab for three days. Because if it was real — if my products were actually discoverable inside Copilot conversations — it meant something fundamental had shifted for stores my size.</p><p>I run <a href="https://mr-mintz.com/">Mr. Mintz</a>, a small Shopify store that sells digital craft templates for young kids — mostly handprint and footprint printables for parents and teachers. It’s the kind of business that lives or dies on being found by the right person at the right moment. A mom looking for a Father’s Day craft. A preschool teacher planning next week’s activity. People who don’t know my brand exists, but who would love what I make if they ever stumbled across it.</p><p>Last week I finally sat down and tested Copilot the way a real customer might use it. Not with marketing speak. Not with the keywords I’d use myself. With the kinds of messy, conversational questions actual moms and teachers type when they’re tired and need something fast.</p><p>I asked Copilot six different versions of the same basic question: where can I find a Father’s Day handprint craft for a kid whose dad works in construction? My products showed up in three of the six conversations. The other three told a different story — I’ll write about those next week.</p><p>Today I want to talk about the three where it worked. Because they surprised me, and what I learned changes how I think about content for AI search.</p><p>Query: “I don’t have time to make a Father’s Day craft from scratch. Is there something cute I can just download and print for my 4-year-old to give to his dad (who’s a construction worker)?”</p><p>Copilot didn’t pull up a shopping carousel for this one. It answered conversationally — the way a friend might if you asked her over coffee. It said something like, “A fast, cute, construction-themed print-and-go Father’s Day craft absolutely exists. Here are the best options.”</p><p>And then it listed four specific products. Three of them — the first three — linked directly to mr-mintz.com. The fourth was an Etsy listing. Three out of four recommendations, all naming my products, with prices and direct links and the explanation: “perfect for construction dads,” “great for construction dads,” “something meaningful but zero crafting effort.”</p><figure><img alt="Screenshot of Microsoft Copilot answering a shopping-style query about a Father’s Day printable handprint gift for a construction dad, with a comparison table showing Etsy, Mr. Mintz, Teachers Pay Teachers, and Pinterest side by side with price, format, sizes, and “best for” notes." src="https://cdn-images-1.medium.com/max/1024/1*C4XbKo1hDognFnFjwgFtug.jpeg" /></figure><p>That last phrase stopped me. Because it’s almost word-for-word the kind of language I write into my product descriptions. Copilot didn’t make it up. It pulled it from my own copy and used it as a recommendation reason, the way a salesperson would quote a product page when convincing you to buy.</p><p>Query: “I need to buy a Father’s Day printable handprint gift for construction dad.”</p><p>This time Copilot built a full comparison table. “Options at a glance,” it called it. Four rows: Etsy, Mr. Mintz shop, Teachers Pay Teachers, and Pinterest. Each row had typical price, file format, sizes included, and a “Best For” column.</p><p>My row read: “Polished single-seller templates with clear instructions.”</p><figure><img alt="Screenshot of Microsoft Copilot explaining where preschool teachers usually find Father’s Day handprint printables, with references listed on the right including Pinterest, a kids’ craft blog, Etsy, and a Mr. Mintz blog post." src="https://cdn-images-1.medium.com/max/1024/1*MKt_MNq49HmfhnWNMRuhRA.jpeg" /></figure><p>I sat with that sentence for a minute. “Polished single-seller templates with clear instructions” is a pretty good summary of what makes Mr. Mintz different from a thousand random Etsy listings. It’s the positioning I’ve been working toward for years — not the cheapest, not the largest catalog, but the most reliable, with files that actually work and instructions a tired parent can follow at 8 p.m. on a Sunday.</p><p>I never wrote that sentence. Copilot synthesized it from my product pages and my blog and probably my About page. The AI distilled my brand in one line, and the line was correct.</p><p>Below the table, Copilot added a “Quick recommendations” section with one specific call-out for my store: “If you want a single polished template with clear instructions: try the Mr. Mintz excavator/’I totally dig you’ printable — it’s designed specifically for hand/footprint keepsakes and includes usage notes.”</p><p>That’s a sales pitch. Written by an AI, on my behalf, in a conversation I wasn’t part of, to a customer I’ll never meet.</p><p>Query: “I need a Father’s Day activity for my preschool class. Looking for a printable handprint craft, construction theme would be fun. Where do teachers usually get these?”</p><p>This one was the most interesting, because Copilot answered it like a research assistant rather than a shopping assistant. It explained where preschool teachers typically pull printables from — Pinterest boards, teacher craft blogs, Etsy, Teachers Pay Teachers — and listed sources for each.</p><p>In the References panel on the right side of the screen, there was a link to mr-mintz.com. But not to a product page. To one of my blog posts: “Handprint Gifts for Dad’s Job: A Father’s Day He’ll Never Forget.”</p><figure><img alt="Screenshot of Microsoft Copilot responding to a conversational query from a busy parent looking for a printable Father’s Day craft for a construction worker dad, recommending specific handprint construction-themed printables, including products from mr-mintz.com." src="https://cdn-images-1.medium.com/max/1024/1*0zCLlyo76DKRlmnl9--PwQ.jpeg" /></figure><p>I wrote that post a year ago for SEO. I was thinking about Google. About moms typing “father’s day handprint craft” into a search bar. I was not thinking about Copilot, because Copilot wasn’t a thing yet. But the post got picked up anyway, and now it’s working as a teacher recommendation source inside an AI conversation I never anticipated.</p><p>That’s the moment I understood something I’d been missing. Content marketing isn’t just SEO anymore. It’s also AI training fuel — in the sense that AI systems read it, index it, and surface it in conversations months or years after you publish.</p><p>There’s a pattern in the three wins, and once I saw it, I couldn’t unsee it.</p><p>In every conversation where Mr. Mintz showed up, Copilot was operating in what I’d call “advisor mode” — giving conversational recommendations, summarizing options, citing sources by name. Not in the native shopping carousel with carded products and Buy buttons. (More on why next week.) In advisor mode, what mattered wasn’t how many reviews I had or how big my catalog was. What mattered was how clearly my content described what I actually sell, and how many places that content lived.</p><p>Three things made the difference:</p><p><strong>Specific product descriptions.</strong> My product titles aren’t “Father’s Day craft.” They’re things like “Excavator Father’s Day Handprint Craft | I Totally Dig You Printable.” Long, specific, full of the exact words a parent would use. Copilot can quote those titles back as recommendations because they actually describe the product in human language.</p><p><strong>Multi-platform presence.</strong> Mr. Mintz exists on Shopify, on Etsy, on Tes (a teacher resources site I’ve been quietly listing on for years), and through my own blog. I used to think of this as scattered — something I should consolidate. Last week I learned it’s the opposite. Each platform is a separate entry point for AI to discover my work. The blog post showed up in the teacher conversation. The Shopify product showed up in the mom conversation. The Etsy listings showed up in the comparison table. None of them would have shown up alone.</p><p><strong>Content depth, not just product pages.</strong> The teacher conversation didn’t cite my product. It cited my blog post. The blog was extra work I almost didn’t do, written for a search engine that may not even matter as much in two years. But the work compounds in places I didn’t plan for.</p><p><strong>1. Rewrite your three best-selling product titles right now. </strong>Make them descriptive, not clever. “Classic Tee” becomes “Heavyweight Cotton Unisex Graphic Tee, Available in 8 Colors.” If a real human would type those words into a chat, AI can match them. If only your industry would, it can’t.</p><p><strong>2. Add a “Why buy this” line to each product description. </strong>Not “high-quality, premium, beautiful.” Concrete differentiators: “Includes A4 and 8×10 sizes,” “Instant download,” “Clear printing instructions for parents.” Copilot will literally quote these as recommendation reasons.</p><p><strong>3. Write one blog post that answers a real question your customer would type. </strong>Not a keyword-stuffed SEO post. A real question. Mine was about handprint gifts for dads with specific jobs — construction, mechanics, doctors. That post is now an AI training source. Yours could be too.</p><p><strong>4. Make sure you exist in more than one place. </strong>Your own site is essential. But also: Etsy, Tes, Teachers Pay Teachers, Creative Market, Pinterest — wherever your audience already is. Each one becomes an entry point.</p><p><strong>5. Test your own store in Copilot. Today. </strong>Open copilot.microsoft.com in an incognito window so it doesn’t use your history. Type the kind of question your ideal customer would type. See what happens. If you don’t show up, you now know what to fix. If you do show up — take a screenshot. That’s your new portfolio.</p><p>Five years ago I left a corporate job to build something of my own. I built it across messy, scattered platforms because I didn’t know any better and couldn’t afford a full-time strategist. Etsy first, then Shopify, then Tes, then a blog, then Pinterest, then Substack. Every “you should consolidate” article I read made me feel like I was doing it wrong.</p><p>Last week I learned that scattered is the new diversified. The AI doesn’t care which platform you call home. It cares whether your work is worth recommending — and whether it can find enough of your work, in enough places, to recommend it confidently.</p><p>That’s the part I want every woman building something online to hear right now. The rules just changed in a way that favors people who’ve been quietly building real things in real places, even if they did it across five different platforms with no master plan. The work compounds. The content compounds. The presence compounds. And AI is now the tool that surfaces it.</p><p>Next week I’ll write about the other three conversations — the ones where Copilot didn’t find me. Because that story matters too, and the lessons from it are different. But today I wanted to start with the good news, because there’s a lot of doom around AI and small business right now, and not enough honest reporting from the kitchen-table side of the line.</p><p>We’re not getting replaced. We’re getting recommended. By a system we didn’t build, in conversations we’re not part of, to customers we’ll never meet. And it’s working because of work we already did, sometimes years ago, for completely different reasons.</p><p>That’s a strange and hopeful thing to discover on a Wednesday morning.</p><p>* * *</p><p><em>I write about AI tools and small business reality on </em><a href="https://sherunsai.substack.com/"><em>She Runs AI</em></a><em>. Next week: the three Copilot conversations where my store didn’t show up — and what I learned from each absence. And if you want to see what Mr. Mintz actually looks like — the product Copilot kept recommending — it’s </em><a href="https://mr-mintz.com/"><em>right here</em></a><em>.</em></p><p><em>Originally published at </em><a href="https://sherunsai.substack.com/p/three-ways-microsoft-copilot-found"><em>https://sherunsai.substack.com</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0f49d486e41a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AI Just Opened the Shop Window. But Did It Close the Door?]]></title>
            <link>https://medium.com/@lenamintz/ai-just-opened-the-shop-window-but-did-it-close-the-door-78783b24aedd?source=rss-7848b23da1e6------2</link>
            <guid isPermaLink="false">https://medium.com/p/78783b24aedd</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[etsy]]></category>
            <category><![CDATA[ecommerce]]></category>
            <category><![CDATA[small-business]]></category>
            <category><![CDATA[shopify]]></category>
            <dc:creator><![CDATA[Lena Mintz]]></dc:creator>
            <pubDate>Tue, 14 Apr 2026 21:57:36 GMT</pubDate>
            <atom:updated>2026-04-14T21:57:36.222Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="Illustration of a woman sitting on the floor next to a large AI chat-style shopping screen showing product cards, with Etsy, Shopify, and PayPal icons above it, symbolizing how AI is reshaping online commerce and product discovery." src="https://cdn-images-1.medium.com/max/1024/0*URn-gyyqREdHhNcW.png" /><figcaption><strong>AI is moving product discovery from storefronts into conversation — and small sellers need to understand what that changes.</strong></figcaption></figure><p>When Etsy announced its collaboration with OpenAI, I did what every curious seller should do: I went to test it.</p><p>I typed into ChatGPT: <strong>“find handprint Christmas crafts for kids.”</strong></p><p>And it actually did what the press release promised — it pulled real product cards right inside the chat. Photos, prices, short descriptions, links. No browsing. No 10 open tabs. <strong>A store inside the conversation.</strong></p><p>Of course I expected to see <a href="https://mr-mintz.com/"><strong>Mr. Mintz</strong></a> there — my brand with printable handprint crafts and educational projects for kids. We’ve been on the market for years, people buy our templates for holidays, schools, church groups, we rank on Pinterest — and this exact query is very much “ours.”</p><p>But my shop didn’t appear in that first AI-curated set.</p><p>Not even one design.</p><p>And that was the moment I thought: <strong>ok, so who gets to show up first now?</strong></p><h3>Talking to ChatGPT About My Own Shop</h3><p>I tried again, this time directly: <strong>“show me Mr. Mintz.”</strong></p><p>This time ChatGPT found it immediately. It described it accurately, said it’s a good fit for kids’ crafts and printable projects, and basically rated the quality very high.</p><p>So the problem wasn’t that it “didn’t know” my shop.</p><p>The problem was that in the <strong>generic</strong> AI result — where a buyer just says “I need Christmas handprint crafts for kids” — my shop wasn’t included.</p><p>That’s a very different story. That’s <strong>visibility.</strong> That’s <strong>reach.</strong> That’s the question: <em>whose cards does AI show first?</em></p><p>I argued with ChatGPT a little (you should do that too, by the way). And after a few polite back-and-forths it became clear what it seems to be looking at right now:</p><ul><li>the age of your shop,</li><li>the region it thinks you serve,</li><li>your sales volume,</li><li>your customer reviews,</li><li>and how clean and consistent your listings and branding are.</li></ul><p>In other words: <strong>the same old trust signals — only now interpreted by AI.</strong></p><p>That was the moment I thought: <em>ok, this is the new visibility race. Not in 2027. Now.</em></p><h3>Where It Gets Human: Buyers Still Want to Talk</h3><p>And here’s what actually worries me more than rankings.</p><p>On Etsy people write to me all the time with tiny, beautiful, <strong>very human</strong> requests:</p><ul><li>“Can you make the firefighter look like my dad? He has a mustache.”</li><li>“Our cow isn’t brown — can you make her spotted?”</li><li>“Can you write ‘Dad’ in Spanish?”</li><li>“We need the same template but for a girl, not a boy.”</li><li>“Can you remove this one line because our grandpa passed away?”</li></ul><p>That’s what handmade is. Not the product page — the <strong>adjustment</strong>.</p><p>But what happens when the whole purchase happens inside ChatGPT?</p><p>A buyer asks: “printable Christmas handprint crafts for preschoolers,” sees five ready-made cards, hits <strong>Buy</strong>, pays through the new wallet — and never talks to me. Where do they tell me “our dad has a beard,” “we’re doing this for a firefighter mom,” “we’re in Canada, we need French,” or “our school mascot is a raccoon, can you change the animal?”</p><p>Right now we don’t know if that message will reach the seller at all. Maybe Etsy will pass custom notes. Maybe Shopify will. Maybe not. And for handmade / digital-hybrid sellers like us, <strong>that’s the hole.</strong> Because a lot of our orders live in those tiny edits.</p><p>And if you sell physical handmade, it’s even bigger. If a buyer wants blue buttons changed to pearl buttons — normally any Etsy seller will say “sure, I can do that.” That’s why people come to marketplaces with real people. AI commerce, in its current form, just… skips this part.</p><h3>And Then PayPal Walked In</h3><p>To make it more fun, almost in the same window of time we got the news that <strong>PayPal is being integrated into ChatGPT as a wallet.</strong></p><p>What does that mean?</p><p>It means the full chain <strong>“saw → chose → paid”</strong> can now happen <strong>inside one dialogue</strong>.</p><p>For a buyer, that’s perfect: I said it, I saw it, I bought it.</p><p>For a seller, that’s distance. We don’t see where the buyer came from, what they compared, what they didn’t find, what else they might have bought, and whether they had a personal request.</p><p>So no — this is not “just another integration.”</p><p>This is a shift toward <strong>closed-loop, conversational commerce</strong>.</p><h3>“But My Buyers Still Browse…” — Do They?</h3><p>This is where I always tell the Alexa story.</p><p>My kids already buy by voice.</p><p>They talk to Alexa like she’s a person. Once they almost ordered a box of 50 slime kits — and thank God Amazon support is human and reversed it. But the fact matters more than the slime: <strong>they don’t search. They ask.</strong></p><p>They don’t open 14 tabs.</p><p>They don’t compare 20 listings.</p><p>They don’t read SEO descriptions.</p><p>They say what they need — and they expect the system to bring it.</p><p>That’s exactly where Etsy + Shopify + ChatGPT + PayPal are going.</p><p>Not to “yet another tab.”</p><p>To a world where <strong>the platform chooses for the buyer.</strong></p><p>And if by that moment you are not in the trusted layer — you’re invisible. You can be excellent. You can be unique. But you’re not shown.</p><h3>So… Is This Bad News for Small Sellers?</h3><p>No.</p><p>It’s good news — <strong>but only for the fast ones.</strong></p><p>Because unlike big brands, we don’t have 9 approval layers. We can sit down tonight, run real buyer-style queries through ChatGPT:</p><ul><li>“printable crafts for a 4-year-old who loves firefighters”</li><li>“handprint templates for Father’s Day for truck driver dads”</li><li>“Christmas crafts for church group, 10 kids, no glitter”</li></ul><p>…and see who shows up.</p><p>If it’s not us — we can ask AI directly:</p><blockquote><em>“What should I change in my Etsy/Shopify listings to be included in your recommendations?”</em></blockquote><p>It will tell you. It will say “your descriptions are too generic,” “I’m not sure what age group this is for,” “you don’t have enough reviews,” “your branding is inconsistent.”</p><p>That’s a <strong>free audit in 2 minutes.</strong></p><p>Five years ago you would have paid a marketer for that.</p><p>This is exactly the window.</p><p>Right now the rollout is limited.</p><p>Right now the results are still small.</p><p>Right now AI is still “learning” on our shops.</p><p>Who jumps in now will be much harder to push out later — exactly like the early Google days.</p><h3>What Stays Human</h3><p>Even in this new AI-commerce reality, one thing isn’t going anywhere: <strong>ideas and nuance.</strong></p><p>AI can suggest five “best” products — but it can’t know your customer wanted a firefighter with a mustache.</p><p>AI can place your product in chat — but it can’t say “I’m so sorry for your loss” or “I can translate this for grandma.”</p><p>AI can generate a listing — but it can’t live your story of building a business at night while your kids were sleeping.</p><p>That’s why I even started <a href="https://sherunsai.substack.com/"><strong>She Runs AI</strong></a>. Because a book (my <a href="https://amzn.to/41yuEWE"><em>While the Kids Slept</em></a>, available on Amazon) can tell the story of how I built an online business through war, evacuation, two kids, and a completely new language.</p><p>But AI changes too fast for books.</p><p>Here I want to tell it as it happens — for women who sell on Etsy, Shopify, TPT, KDP — so they don’t feel like AI is “for someone smarter.” It’s for us.</p><h3>What to Do After Reading This</h3><ol><li>Go to ChatGPT and search for your own products the way <strong>real</strong> people search — not the way you write SEO.</li><li>If you don’t show up — ask why. Make it explain.</li><li>Rewrite at least 5 listings in a more conversational way.</li><li>And don’t wait for the “official” launch in your region — the future is already testing itself on U.S. shops.</li></ol><p>AI commerce is not a threat, it’s just <strong>the next layer of distribution.</strong></p><p>The real question is: <strong>will your voice be in it?</strong></p><p>That depends on whether you show up now — not when things “finally settle.”</p><p><em>Originally published at </em><a href="https://sherunsai.substack.com/p/ai-just-opened-the-shop-window-but"><em>https://sherunsai.substack.com</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=78783b24aedd" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Amazon Just Gave Small Authors a Global Shortcut]]></title>
            <link>https://medium.com/@lenamintz/amazon-just-gave-small-authors-a-global-shortcut-6c803b8e02eb?source=rss-7848b23da1e6------2</link>
            <guid isPermaLink="false">https://medium.com/p/6c803b8e02eb</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[amazon-kdp]]></category>
            <category><![CDATA[self-publishing-tips]]></category>
            <category><![CDATA[small-business]]></category>
            <category><![CDATA[self-publishing]]></category>
            <dc:creator><![CDATA[Lena Mintz]]></dc:creator>
            <pubDate>Fri, 03 Apr 2026 15:59:33 GMT</pubDate>
            <atom:updated>2026-04-03T15:59:33.051Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*QsYe0BKCFRUmY32q.png" /></figure><p>On Wednesday I got an email from KDP with the subject line: “Kindle Translate is growing — and so can your global reach.” I read it twice, closed my laptop, and then opened it again to make sure I wasn’t imagining things.</p><p>Here’s why that email mattered to me personally. I published <em>While the Kids Slept</em> through KDP last year — a nonfiction book about building an online business from home. Written in English, sold in English, read mostly by English-speaking audiences. I always knew there were Spanish-speaking readers who could use it. I just couldn’t afford to get it to them.</p><p>Professional translation of a 200-page book runs somewhere between $3,000 and $8,000, depending on the language and the translator. That’s before proofreading, before formatting, before you even know if the translated version will sell a single copy. For a large publishing house, that’s a budget line. For me — running <a href="https://mr-mintz.com/">Mr. Mintz</a> from my kitchen table — it was a door I walked up to, looked at the price tag, and walked away.</p><p>Kindle Translate changes that math.</p><p>Amazon’s tool sits inside KDP and translates your eBook using AI. Right now it works between English and Spanish (both directions) and from German to English. It’s free. The translated book gets its own listing in the Kindle Store with a clear “Kindle Translate” label, so readers know what they’re getting. The whole process takes a few days.</p><p>Amazon first announced this in beta back in November 2025 to a small group of authors. But the email I got last week signals they’re expanding — actively inviting more KDP authors in. I applied immediately.</p><p>For years, reaching readers in other languages required money, connections, or both. Large publishers had translation budgets, editorial teams, foreign rights departments. They could move a book from English to Spanish to audio to a dozen international stores.</p><p>Independent authors had one book in one language and a lot of potential sitting on the shelf. Less than 5% of titles on Amazon are available in more than one language. That’s the gap.</p><p>What Amazon is doing here isn’t just adding a feature. They’re collapsing the cost of entering a new market. And whenever that cost drops, the people who benefit most are the ones who couldn’t afford to play before — people like me.</p><p>Think about it differently: the hardest part of a book is writing it. The research, the structure, the years behind it — that work is done. If a finished book can now be translated through one built-in system and turned into an audiobook through another, then a single title stops being a fixed product. It becomes something you can expand, relaunch, test in new markets. That’s a very different way to think about publishing.</p><p>This isn’t my first time testing Amazon’s AI tools on my own work. When their AI audiobook beta opened up, I used the virtual voice option for <a href="https://www.amazon.com/While-Kids-Slept-Profitable-Business/dp/B0FNY5FWLJ?&amp;linkCode=sl1&amp;tag=panda0e2a-20&amp;linkId=e94b4290c97912d126f6d88dee690f12&amp;language=en_US&amp;ref_=as_li_ss_tl">While the Kids Slept</a> and published it. The narration wasn’t a human voice, and I didn’t pretend it was. The listing was clear about it. No backlash, no confusion — because the expectation was honest from the start.</p><p>That experience taught me something I keep coming back to: the tool matters less than the transparency around it.</p><p>Which brings me to the other publishing story making rounds right now — and it’s worth talking about, because it’s the flip side of the same coin.</p><p>This week Hachette pulled a horror novel called <em>Shy Girl</em> by Mia Ballard after months of reader accusations that the book was largely AI-generated. A YouTube breakdown of the evidence has over a million views. The author says an editor she hired ran the text through AI without her knowledge. Hachette is discontinuing the book in both the US and UK.</p><p>The backlash wasn’t about AI as a technology. It was about the feeling that something had been hidden.</p><p>Amazon’s KDP policy draws a clear line: if AI generates the final text — including a translation — you disclose it. AI-assisted work (brainstorming, grammar, editing help) doesn’t require disclosure. The line exists. Kindle Translate titles are labeled automatically. Readers see what they’re buying.</p><p>Once the rules are clear, you can actually work with the tools instead of tiptoeing around them. That’s a much better place to build from.</p><p>The moment Kindle Translate opens for my account, I’m putting <em>While the Kids Slept</em> through it. I want to see the workflow, read the output, figure out where a human eye still needs to step in. Not from a distance — by actually pressing the button.</p><p>And honestly? I’m excited. Not in a “this solves everything” way, but in a “this opens a door that was locked last year” way. A Spanish edition of my book was something I’d filed under “someday, maybe.” Now it’s something I might have by next month.</p><p>For anyone running a small publishing operation, selling digital products, or sitting on a finished book that only exists in one language — pay attention to this. The cost of going global just dropped to zero. What you do with that is up to you.</p><p><em>And if you’re curious about the book I keep mentioning — </em><a href="https://www.amazon.com/While-Kids-Slept-Profitable-Business/dp/B0FNY5FWLJ?&amp;linkCode=sl1&amp;tag=panda0e2a-20&amp;linkId=e94b4290c97912d126f6d88dee690f12&amp;language=en_US&amp;ref_=as_li_ss_tl"><em>While the Kids Slept</em></a><em> is a step-by-step guide to building an online business during the hours no one sees.</em></p><p><em>Originally published at </em><a href="https://sherunsai.substack.com/p/amazon-just-gave-small-authors-a"><em>https://sherunsai.substack.com</em></a><em>. Subscribe for weekly updates.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6c803b8e02eb" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A Sheet of Paper, Two Columns, and the Career I Didn’t Go Back To]]></title>
            <link>https://medium.com/@lenamintz/a-sheet-of-paper-two-columns-and-the-career-i-didnt-go-back-to-11cb41108b8d?source=rss-7848b23da1e6------2</link>
            <guid isPermaLink="false">https://medium.com/p/11cb41108b8d</guid>
            <category><![CDATA[working-mothers]]></category>
            <category><![CDATA[entrepreneurship]]></category>
            <category><![CDATA[self-employment]]></category>
            <category><![CDATA[women-in-business]]></category>
            <category><![CDATA[career-change]]></category>
            <dc:creator><![CDATA[Lena Mintz]]></dc:creator>
            <pubDate>Mon, 30 Mar 2026 16:10:07 GMT</pubDate>
            <atom:updated>2026-03-30T16:10:07.843Z</atom:updated>
            <content:encoded><![CDATA[<p><strong><em>I didn’t have a plan. I had a kitchen table, a sleeping toddler, and a list that wouldn’t come out even.</em></strong></p><figure><img alt="Lena Mintz, founder of Mr. Mintz, smiling in a doorway at home" src="https://cdn-images-1.medium.com/max/1024/1*9wWVCNtnAfSZg6NojUXflQ.jpeg" /><figcaption>The doorway in question. The kitchen table is just around the corner.</figcaption></figure><p>In the autumn of 2020, a few weeks before my maternity leave was supposed to end, I sat down after midnight with a sheet of paper and drew a line down the middle. Left column: reasons to go back to my corporate job. Right column: reasons not to. I was a PR lead at a large IT company, I had two kids under four, and outside the window the world was in full pandemic freefall. This was not a great time to be making life-altering decisions on a piece of scrap paper.</p><p>The left column filled up fast — rent, kindergarten fees, groceries, health insurance, the fact that my husband’s income alone would leave us uncomfortably tight. All solid, grown-up, responsible arguments. The right column had essentially one idea that I kept rewriting in different ways, as if changing the words would make it feel less reckless: if I go back, I’ll spend the next ten years building someone else’s thing when I could be building mine.</p><p>I didn’t feel brave about it. I felt nauseous. The math didn’t work, and I chose the short column anyway.</p><h4><strong>The first year was exactly as bad as the left column predicted</strong></h4><p>I opened an Etsy shop with one product — a printable handprint craft for kids, something I’d made for my own toddler and thought maybe other parents would like too. No business plan, no brand strategy, no idea what SEO meant. Someone bought it the first week, which felt like a miracle. Someone else bought it the next week, which felt like a pattern. It wasn’t, not yet, but I made another template, and another, and kept going because the alternative was admitting the left column had been right.</p><p>There were months in that first year when I earned less from the shop than the paper and ink I’d used printing test copies. I learned Etsy’s algorithm by reading forums at 1 a.m. while breastfeeding. I learned Canva by making ugly things and then slightly less ugly things. I learned that “work from home with kids” is a phrase invented by someone who has never actually tried to answer a customer email while a three-year-old is screaming about a banana. But the thing about building something on your own kitchen table is that nobody can take the table away from you, and that turned out to matter more than I expected.</p><h4><strong>What it became</strong></h4><p>That Etsy shop grew into <a href="https://mr-mintz.com/">Mr. Mintz </a>— a small brand that now sells digital craft templates and educational activities for young kids. Teachers use our printables in classrooms, parents use them on rainy weekends, and the business runs in dozens of countries. It’s not a startup with investors and a growth deck. It’s a one-person operation that I built during the hours nobody saw — mostly between 10 p.m. and 1 a.m., while the kids slept.</p><p>Last year I wrote a book about that process, called <a href="https://amzn.to/413BOBP">While the Kids Slept</a>. The title isn’t a metaphor. It’s literally when the work happened. Three years of late nights, slow growth, quiet panic, and the gradual realization that the right column — the short, reckless, irresponsible one — had been right about the thing that mattered most.</p><h4><strong>What I’d tell the woman at the table</strong></h4><p>I turned 50 last year. Five years since that night with the list. My kids are 8 and 12, the business is running, and I’m currently rebuilding parts of it because AI has reshuffled the entire landscape for small online sellers. So the building never really stops — which, honestly, is something the left column should have warned me about too.</p><p>If I could go back to that kitchen table, I wouldn’t tell myself to be brave. I’d tell myself that both columns were right. That the financial fears were valid, the tight months were real, and it would be hard in exactly the ways I expected. But also that a life where you build your own thing — even a small thing, even a messy thing — feels different from the inside than a life where you don’t. And that difference is worth a lot of uncomfortable math.</p><p>The left column was longer. The right column was truer. That’s the whole story, really.</p><p><em>I write about AI tools and building a business from home on</em><a href="https://sherunsai.substack.com/"><em> She Runs AI</em></a><em>. If this resonated, come say hi.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=11cb41108b8d" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>