<?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 Nassuf Mmadi on Medium]]></title>
        <description><![CDATA[Stories by Nassuf Mmadi on Medium]]></description>
        <link>https://medium.com/@nassuf?source=rss-309baa13bdc3------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*oIHaoTKs8Ja4hVHaKpru3A@2x.jpeg</url>
            <title>Stories by Nassuf Mmadi on Medium</title>
            <link>https://medium.com/@nassuf?source=rss-309baa13bdc3------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 08:02:10 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@nassuf/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[Easy fast social media content for your SaaS ?]]></title>
            <link>https://medium.com/@nassuf/easy-fast-social-media-content-for-your-saas-b96d7443eba2?source=rss-309baa13bdc3------2</link>
            <guid isPermaLink="false">https://medium.com/p/b96d7443eba2</guid>
            <category><![CDATA[solopreneur]]></category>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[saas]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[marketing]]></category>
            <dc:creator><![CDATA[Nassuf Mmadi]]></dc:creator>
            <pubDate>Sun, 07 Jan 2024 16:18:54 GMT</pubDate>
            <atom:updated>2024-01-07T16:18:54.756Z</atom:updated>
            <content:encoded><![CDATA[<h3>Easy fast social media content for your SaaS ?</h3><p>This is not a proper article but just a sharing of an interesting video I recently discovered on YouTube.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FlqXWVzWkkyc%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DlqXWVzWkkyc&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FlqXWVzWkkyc%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/e3c8969d932ec470c496b0e823fc4bdc/href">https://medium.com/media/e3c8969d932ec470c496b0e823fc4bdc/href</a></iframe><p>Basically, you can generate hundreds of shorts / Tiktok posts in 3 steps that takes less than an hour.</p><ol><li>Create a video template using <a href="https://www.canva.com/">Canva</a></li><li>Generate the content of each video with ChatGPT</li><li>Use the “bulk create” app from Canva to automate the video creation</li></ol><p>I have then seen many videos like this talking about how to monetize these shorts.</p><p>I don’t know about this kind of monetization but this technique sounded like an interesting way to create “social medium content” easily for a SaaS tool.</p><p>I am <a href="https://www.tiktok.com/@ppcassist">currently trying it</a>, with 45 Tiktok reels that will be published over the next 90 days. <br>Too soon to say if it helps to get followers &amp; users but I thought it worth sharing.</p><p>Have any of you guys tried this kind of approach for content creation/acquisition?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OFGQUhOr1II075CP5_zcrg.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b96d7443eba2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A personal ghostwriter with GPTs]]></title>
            <link>https://medium.com/@nassuf/a-personal-ghostwriter-with-gpts-e817ac958916?source=rss-309baa13bdc3------2</link>
            <guid isPermaLink="false">https://medium.com/p/e817ac958916</guid>
            <category><![CDATA[saas]]></category>
            <category><![CDATA[marketing]]></category>
            <category><![CDATA[development]]></category>
            <category><![CDATA[gpt]]></category>
            <category><![CDATA[openai]]></category>
            <dc:creator><![CDATA[Nassuf Mmadi]]></dc:creator>
            <pubDate>Tue, 02 Jan 2024 02:53:00 GMT</pubDate>
            <atom:updated>2024-01-02T02:53:00.464Z</atom:updated>
            <content:encoded><![CDATA[<p>One of my goals for 2024 is to create a blog for my SaaS.<br>But as you know, time is limited. <br>Especially when you are a solo founder on a bootstrapped project.</p><p>Months ago, I thought about using ChatGPT to write articles, gave it a try, but the results were not “human” enough, so I gave up on that.</p><p>OpenAI has evolved a lot since.</p><p>They might be a great solution to get articles written in our own writing style, fast…</p><p>Here is what I tried and the first results.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Ftenor.com%2Fembed%2F13800248&amp;display_name=Tenor&amp;url=https%3A%2F%2Ftenor.com%2Ffr%2Fview%2Fsnoop-dogg-dre-riding-driving-rolling-gif-13800248&amp;image=https%3A%2F%2Fmedia.tenor.com%2Fd6Nqngv_DnsAAAAF%2Fsnoop-dogg-dre.png&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=tenor" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/91d60fcd9e5a8c489257b271268bcba0/href">https://medium.com/media/91d60fcd9e5a8c489257b271268bcba0/href</a></iframe><h3>Dr. Dre, Jay-Z + Microconf + BannerBear + SRICS = 💡</h3><p>I just spent the last week slowing down on <a href="https://ppcassist.com/">PPCAssist</a> and enjoying great bootstrapped SaaS-related content:</p><ul><li>The audiobook “<a href="https://www.amazon.com/SaaS-Playbook-Multimillion-Dollar-Startup-Without/dp/B0CDZHLFL8">The SaaS Playbook</a>” by Rob Walling. <br>Full of practical advice that helps if you are trying or willing to bootstrap a SaaS.</li><li>A lot of videos and experience sharing from his great YouTube channel <a href="https://www.youtube.com/@MicroConf">Microconf</a></li><li>Some <a href="https://www.bannerbear.com/open/">very insightful articles</a> and <a href="https://www.indiehackers.com/post/my-bootstrapped-saas-hit-22k-mrr-ama-a0c486b898">indiehacker</a> posts from <a href="https://twitter.com/yongfook">Jon Yongfook</a>, founder of <a href="https://www.bannerbear.com/">BannerBear</a> about what he is doing at each step of his business. This is gold!</li></ul><p>Besides that, I learned an interesting fact in a <a href="https://youtu.be/uv8fiajMtCA?si=WHjXWtQ2AMZj_-6M&amp;t=512">French YT video</a> about Dr. Dre, Snoop Dogg and Jay-Z:</p><p>One of the most known Dr. Dre’s tracks, <a href="https://www.youtube.com/watch?v=_CL6n0FJZpk">“Still”</a>, was entirely written by a ghostwriter… Jay-Z!</p><p>The new year came (Happy New Year by the way 🥳) and my SaaS chat started to ring again. Questions, bugs, back to stress-increasing notifications 🫣</p><p>Back to business.<br>I did a S.R.I.C.S, aka Stress-Relief-Indoor-Cycling-Session 😅<br>As usual, with stress relief, comes fresh ideas mixing what I had seen and heard recently:</p><blockquote>“What if I create my own ghostwritter using GPTs to write blog articles related to Amazon Ads ?<br>I could inject all my medium articles in a personal GPT and instruct it to use my writing style ?”</blockquote><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Ftenor.com%2Fembed%2F7178836&amp;display_name=Tenor&amp;url=https%3A%2F%2Ftenor.com%2Ffr%2Fview%2Fbravo-idea-guiness-gif-7178836&amp;image=https%3A%2F%2Fmedia.tenor.com%2FrFRivt4UvcQAAAAF%2Fbravo-idea.png&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=tenor" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/14f4f029766b33aa14af6380853f24ce/href">https://medium.com/media/14f4f029766b33aa14af6380853f24ce/href</a></iframe><p>GPTs were introduced <a href="https://openai.com/blog/introducing-gpts">in November</a> 2023.<br>Watching <a href="https://www.youtube.com/watch?v=U9mJuUkhUzk">Sam Altman’s keynote</a> back then, one new feature blew my mind and seems, like a key to building incredible solutions!</p><p>The ability to upload content and set a purpose to the GPT.</p><p>At that time I thought about using that on my SaaS to help sellers analyze their ads faster by just discussing with an assistant.<br>I gave it a try. <br>Created a GPT, uploaded some test ads data, and tried several prompts. The results were mixed. <br>My conclusion was that I needed to invest more time in a POC to overcome GPT limitations.<br>I will do that in the next weeks and let you guys know about it.</p><p>But what about ghostwriting?<br>On paper, it sounds like a great way to produce valuable SEO content fast, with a human-like writing style.</p><h3>POC</h3><p>Setting up the test was very easy (️️❤️OpenAI)!</p><p>I used OpenAI Assistant but you can also do it in ChatGPT by going to <a href="https://chat.openai.com/gpts/discovery">GPTs</a> and then “Create a new GPT”. I am not sure it will use the same GPT model tho.</p><p>So:</p><ol><li>I downloaded all my medium articles into Word .doc files</li><li>Uploaded them to the assistant</li><li>Define these instructions<br>“You are Nassuf Mmadi’s ghostwriter.<br>Nassuf is a software engineer, ex Amazon seller, who build PPCAssist, a SaaS that helps Amazon sellers with their Amazon Ads (PPC) by letting them apply a strategy to receive optimization suggestions.<br>In the attached files you will find articles written by Nassuf Mmadi. Your goal is to write blog article for PPCAssist using the same style as Nassuf”</li></ol><p>In less than 15mn (the time to manually copy and save all articles) it was ready!</p><p>Here is the first result:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KQOfJLr-HSvJ2Ue_TQLBmw.png" /></figure><p>Except for the quote of “Cyril Bentz”, one of the Amazon ads expert I am working with and that I mentioned in an article, it didn&#39;t look like it uses my writing style so I asked for a more personalized result.</p><p>Here it is:</p><p><strong>User</strong></p><blockquote>Rewrite the article using Nassuf Mmadi’s style based on the attached articles</blockquote><p><strong>Nas Ghost Writer</strong></p><blockquote>Hey Amazon sellers, Nassuf here to demystify PPC strategies for 2024, focusing on how to tackle that pesky high ACOS head-on!</blockquote><blockquote>PPC, our beloved pay-per-click advertising on Amazon, isn’t just a good-to-have anymore — it’s your golden ticket to the competitive leagues. But with great power comes, well, the great responsibility to use it right. If you’ve been around since ’22, you’d remember a time when organic spots were the talk of the town. Fast forward, and it’s clear as day: you’ve gotta play the PPC game to win.</blockquote><blockquote>Now, I’ve seen some folks run their PPC like they’re jogging on a treadmill — steady, without much thought, just burning through cash without gaining traction. But here’s the game-changer: PPC should be a stepping stone towards organic glory. Imagine this: you target that perfect keyword, make a few initial PPC-driven sales, and over time, these convert into organic sales, boosting your ranking and — voilà! — you’re making sales with less ad spend.</blockquote><blockquote>Let’s zero in on real strategy, one that Cyril Bentz nails down pretty neatly:</blockquote><blockquote>Pick your keyword.</blockquote><blockquote>Keep an eagle eye on your CTR and conversion rate.</blockquote><blockquote>Aim for a conversion rate that’s soaring above your store’s average.</blockquote><blockquote>Watch your organic rank like you’re binge-watching your favorite series; if it climbs, you’re on the right track.</blockquote><blockquote>The catch is, monitoring all this manually can be a real grind, especially if you’re juggling a bunch of keywords and products. It’s enough to make your head spin.</blockquote><blockquote>Enter the realm of PPC automation. These software promise to take the weight off your shoulders, adjusting bids and chasing those top-performing keywords using all sorts of algorithms. But, let me be clear, if your business isn’t cranking out tens of thousands of orders every month, those fancy AI claims might just be smoke and mirrors.</blockquote><blockquote>Here’s a shocker: I’ve seen an AI spend 380 euros on dud keywords in less than a week, on an account that’s bringing in over 50k a month. Not cool.</blockquote><blockquote>So, what’s the sweet spot? How about a tool that’s like having a PPC wingman? Yes, that’s PPCAssist for you. It’s all about striking that beautiful balance between smart automation and your savvy human decision-making. Our tool is about empowering you — the seller — with top-notch insights and advice, all while keeping you updated on how your biz is shaping up, week by week.</blockquote><blockquote>With PPCAssist, say goodbye to flying solo and hello to a co-pilot that guides you with human intelligence-powered automation.</blockquote><blockquote>So, as we plunge into 2024, keep this in your back pocket: PPC isn’t just about spending money; it’s about spending it smartly. Make every penny count, keep that ACOS down, and let PPCAssist be the key to your Amazon success story!</blockquote><blockquote>And because I’m all about that Christmas spirit, I’m throwing out invites to our beta — 30 spots, all on the house. Get on the list, and let’s take your PPC game to the stratosphere.</blockquote><blockquote>Until next time, keep optimizing, and remain profitable! 🎁 Merry Christmas, folks! 🎄</blockquote><blockquote>Nassuf Mmadi</blockquote><p>I think this time it looks more like my kind of writing. Of course this is just a starting point and the final result will always need validation and adjustments.</p><p>With prompt adjustments, it can be a great start to producing SEO content fast with a very personal tone.</p><p>I only tried 2 prompts so far and it took me a few minutes so I guess trying more approaches and prompts would lead to incredible results.</p><p>That said, I think the best ideas/solutions come by crossing other people’s ideas and experiences so I would be glad to read your thoughts, do a SRICS 😅, before continuing on this experiment and sharing about it.</p><h3>Let&#39;s try together and share</h3><p>What are your thoughts?<br>Have you tried GPTs to write “personalized” content to help your business?<br>Would be great to gather experiences in the comment section to help each other out on this long and lonely bootstrapped SaaS path 💪🏿</p><p>Share your thoughts, experiences, and ideas 🙌</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Ftenor.com%2Fembed%2F15495590&amp;display_name=Tenor&amp;url=https%3A%2F%2Ftenor.com%2Ffr%2Fview%2Fhazel-english-shaking-polyvinyl-dance-dancing-gif-15495590&amp;image=https%3A%2F%2Fmedia.tenor.com%2FDblxciHQGlYAAAAF%2Fhazel-english-shaking.png&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=tenor" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/4100612ccf2234a94ab0b4495e797181/href">https://medium.com/media/4100612ccf2234a94ab0b4495e797181/href</a></iframe><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*Sl3OvMnVrDBro1u0NXF2aQ.jpeg" /><figcaption>Jay-ZPT 👀🏃‍♂️</figcaption></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e817ac958916" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Products I love… or hate.]]></title>
            <link>https://medium.com/@nassuf/products-i-love-or-hate-f6e9ee8054cd?source=rss-309baa13bdc3------2</link>
            <guid isPermaLink="false">https://medium.com/p/f6e9ee8054cd</guid>
            <category><![CDATA[laravel]]></category>
            <category><![CDATA[saas]]></category>
            <category><![CDATA[bootstrapping]]></category>
            <category><![CDATA[solopreneur]]></category>
            <category><![CDATA[developer]]></category>
            <dc:creator><![CDATA[Nassuf Mmadi]]></dc:creator>
            <pubDate>Sun, 22 Oct 2023 20:10:59 GMT</pubDate>
            <atom:updated>2023-10-22T22:56:21.858Z</atom:updated>
            <content:encoded><![CDATA[<p>Working on a bootstrapped SaaS led me to discover a lot of tools. I tested, liked, hated, and loved multiple ones. <br>Here are eight of those I love… or hate.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FR6gvnAxj2ISzJdbA63%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2FR6gvnAxj2ISzJdbA63%2Fgiphy-downsized-large.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2FR6gvnAxj2ISzJdbA63%2Fgiphy-downsized-large.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="435" frameborder="0" scrolling="no"><a href="https://medium.com/media/21b0ecac314503a5be41325a1179d25a/href">https://medium.com/media/21b0ecac314503a5be41325a1179d25a/href</a></iframe><h3>🚀 Fast deployment thanks to Laravel Forge</h3><p>When I started, I rented an OVH virtual server that I had to configure myself. I needed to manually install everything (PHP, Postgres SQL, Redis etc) and did every deployment manually.</p><p>As I discovered Laravel and its ecosystem I gave a try to <a href="https://forge.laravel.com/">Laravel Forge</a>.</p><p>Forge manages all configuration and deployment parts for you. I just had to take a Digital Ocean server, set the credentials on Forge, and voilà.</p><p>Laravel Forge manages deployment in one click. Can automatically deploy when you merge on the production branch. Handle supervisord to make sure your queues are always running.</p><p>Simple, easy to use, and time-saving. One love for Laravel Forge ❤️</p><h3>💄 Easy pretty UI with Preline UI</h3><p>If you are using TailwindCSS you will love these UI components.</p><p>I am very bad at design UI. I think I can win an “ugly interface” contest without much effort 😅. <br>I moved from Bootstrap CSS to Tailwind CSS when I started this SaaS. I tried a lot of solutions to get quick and nice user interfaces.</p><p>Most of them were either paying or incomplete.</p><p>A friend of mine (Think-Learn-Man) told me about Preline.<br>Now I use it every time I need a UI component.<br>❤️ <a href="https://preline.co/">PrelineUI</a> is free!</p><h3>🫣 Laravel Spark…</h3><p>A SaaS needs a payment system and a very small team (2 developers) needs to be fast. That is why I chose Laravel Spark to quickly build the payment / billing system.</p><p>On paper it is perfect:</p><ul><li>One time payment</li><li>Manages bills</li><li>Connects with stripe</li><li>Quick to setup</li></ul><p>In our case, we have 3 plans depending on the user’s revenue. We needed to display only the right one: impossible without tricks. I ended up struggling with it and losing time instead of being more efficient.</p><p>Laravel Spark is the only Laravel ecosystem tool that I hate… so far.<br>Here are <a href="https://medium.com/@nassuf/laravel-spark-is-crap-dont-use-it-e24d62e7737e">full details about my Laravel Spark experience.</a></p><h3>🗣️ Feedback &amp; public roadmap with Sleekplan</h3><p>Sleekplan is one of those tools that matches exactly what you need at the moment you need it. It is simply a masterclass in terms of execution!</p><p>Getting user feedback is crucial to improve the tool. I used to ask them for feedback via our Discord server. But the more the community was growing, the less we got feedback.</p><p>I thought it was maybe because they were not comfortable sharing their opinion publicly so I looked for a feedback module. That is when I discovered <a href="https://sleekplan.com/">Sleekplan</a> 😍</p><p>Very smooth onboarding that answered some questions I had, quick and easy setup, and free trial (without credit card). I got used to it very quickly and managed to get way more feedback, centralize them, and even have a public roadmap.</p><h3>💬 Chat with clients with Drift</h3><p>I want to be very responsive with users so a chat module is absolutely necessary.</p><p>A lot of websites use Intercom so that is the first solution I checked. I didn’t even try it, as its pricing starts at $74/month.</p><p>After some quick searches, I discovered <a href="https://www.drift.com/">Drift</a>, tried it and still use it until now.</p><ol><li>They have a free plan… ❌ Live edit, I just checked back and it seems that they don&#39;t provide a free plan anymore. I am still using it for free (one seat). Let me know if you find a way to register for the free live chat module.</li><li>A mobile app to receive your user’s client messages directly in your phone and even a macOS app.</li><li>You can add metadata to logged in users so you know who you are talking with. In my case I added the firstname, company name and onboarding step.</li></ol><p>The only issue (if we can consider that as an issue…) is that it doesn&#39;t unify the user&#39;s feed. You end up with multiple feeds for the same user.</p><h3>✉️ Mailjet</h3><p>With +11k email sent so far, Mailjet is the solution I used for transaction and automation email.<br>You can use public template or create your own one using HTML. <br>They have a free plan that includes all basic features (templates, sending via API etc) as long as you send less than 200 emails per day.</p><p>I moved to the paid plan to use their automation module. It lets you send email sequences when an event occurs.</p><p>In my case, when the users start their trial it starts a 5-emails onboarding sequence over a 14-day period.</p><h3>👀 Good old Sentry</h3><p>I could not do without it! <br>While building a saas, or any kind of software, you will obviously have bugs and errors. <a href="https://www.sentry.io/">Sentry</a> helps me track that easily. I even see the slow pages and slow queries. A must-have!</p><blockquote>Little nice tip.<br>I added user context in a middleware so sentry is aware of the current user. Anytime I get a blocking error in production (500 error) I receive a notification with the error info and the user email and name.<br>This way I can quickly fix it and send a personnalized message to the user.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wgKWsU6mRU6XTVcpuRUrTA.png" /></figure><h3>🤖 Make automation</h3><p>I just talked about receiving error notifications from Sentry.<br>Well, that is done thanks to <a href="https://www.make.com/">Make.com</a></p><p>Make is a <a href="https://zapier.com/">Zapier</a> alternative. I use it to connect tools and create useful workflows like:</p><ul><li>Sending a mobile push notification + a discord message when there is a new Sentry issue</li><li>Create a new card in Trello when new feedback is submitted in Sleekplan</li><li>Receive nice notifications when I get a new subscriber on Stripe</li></ul><p>I also developed a <a href="https://www.make.com/en/integrations/datamz">make connector for Amazon sellers</a> that lets them connect their data to any of the +1400 apps included with Make.</p><h3>🫵 What about you?</h3><p>I am really curious to discover new tools used by fellow SaaS makers. <br>Tools like Sleekplan helped me with feedback but I also learned a lot from the way it is executed. <br>Its landing page, onboarding, and SEO content are brilliant, knowing that it is a solo-founded tool.</p><p><strong>What are you using that helps you save time and build faster?</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f6e9ee8054cd" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The reality behind bootstrapping a SaaS • Part 2 • Pivot deadlock]]></title>
            <link>https://medium.com/@nassuf/the-bootstrapped-saas-journey-part-2-pivot-deadlock-cf38dae7d1e1?source=rss-309baa13bdc3------2</link>
            <guid isPermaLink="false">https://medium.com/p/cf38dae7d1e1</guid>
            <category><![CDATA[bootstrapping]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[amazon-fba]]></category>
            <category><![CDATA[solopreneur]]></category>
            <category><![CDATA[saas]]></category>
            <dc:creator><![CDATA[Nassuf Mmadi]]></dc:creator>
            <pubDate>Thu, 07 Sep 2023 01:48:57 GMT</pubDate>
            <atom:updated>2023-09-07T15:03:43.694Z</atom:updated>
            <content:encoded><![CDATA[<p>In the <a href="https://medium.com/@nassuf/growing-a-bootstrapped-saas-is-so-hard-d89f9990eb7f">previous article</a>, I talked about how I started <a href="https://datamz.me/">datamz</a>, a SaaS for Amazon sellers, and how I came up with the idea of pivoting into targeting Amazon experts instead of sellers.</p><p>Experts were easier to target on LinkedIn. Most of the ones I talked with struggle with Amazon reports and use Excel for reporting. Some software solutions already existed but were too expensive for them.</p><p>So I changed my targets. datamz would now provide “quick and easy reporting for Amazon experts”</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FfxwrhUeEyFkAKgwxqr%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2FfxwrhUeEyFkAKgwxqr%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2FfxwrhUeEyFkAKgwxqr%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="244" frameborder="0" scrolling="no"><a href="https://medium.com/media/27492a82a97f782b00a1658b0cc83d58/href">https://medium.com/media/27492a82a97f782b00a1658b0cc83d58/href</a></iframe><p>I started targeting Amazon experts.<br>I first created a LinkedIn group to “attract” them.<br>I changed the landing page and value proposition. <br>It was all about how fast and easy Amazon reporting is with datamz. <br>I put some nice dashboard screenshots and added a weekly email reporting for those who do it manually every week for their clients.</p><p>From there it was time to market and sell!</p><p>I used <a href="https://www.waalaxy.com/">Waalaxy</a>, a tool that automate LinkedIn invitations and messages, to target thousands of Amazon experts.<br>Most of them are in Pakistan. I started understanding their ecosystem:</p><ol><li>The majority of them learned about Amazon and PPC at <a href="https://www.enablers.org/">Enablers</a>, the largest Pakistani e-commerce network/courses providers.</li><li>Then they struggle to find clients and all have the same strategy: Facebook groups, Upwork, and LinkedIn, posting the same Amazon ads console screenshots to prove their results.</li><li>The ones who finally get clients spend hours per week extracting Google Sheets reports for analysis and client reporting.</li></ol><p>I sent thousands of invites, got hundreds of new connections, talked to +300 experts, and called +100.</p><p>I got a lot of “sounds great, I will register this week” followed by ghosting 👀</p><p>Finally succeeded in onboarding only around 10 of them.</p><p>Even if they found the dashboard and reporting useful, none of them were ready to pay for it. They all told me they need to convince their client to pay for the tool but won’t pay themself even if it saves time.<br>I came to understand that they charge $150 to $400 per month, sometimes less. Paying $20–30 for a tool to save time is not relevant in that case. I should have started by asking that…</p><p>So I decided to reached out to all French experts. Literally all of them! <br>I found them through LinkedIn and a <a href="https://malt.fr/">french freelance website</a>. Met and talked with some nice people, onboarded 4.<br>One of them was an agency managing 13 stores.</p><p>They were two girls. They told me it took them 1 hour every week to prepare their reporting to their client.<br>They showed me their spreadsheet reporting and were very excited about the idea of having the same reporting instantly and always up-to-date.<br>I started working on it !</p><p>One week later the reporting page was ready! <br>They could have the metrics they want, and filter on any selling country and period!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*n5h68qypIpGMCSTo1rNjEA.png" /></figure><p>But an unexpected issue arose: the data was not accurate!</p><p>The Amazon Selling-Partner API provides different sources of data for the same metrics. <br>Depending on which report you synchronize, the page views, sessions, and even revenue can differ. <br>It took days of adjustments. Back and forth. Each time I changed something, they had to check and validate it.</p><p>During this period, I had no way to determine if the few beta users were using the tool, and if so, what they were using. So, I began adding some logs and a lightweight admin dashboard where I could track user activity.</p><p>I continued to change the report synchronization sources and asked them to compare it with the actual data. Eventually, they stopped responding and ceased using the tool altogether, including the main dashboard. 🏃🏻‍♀️️🏃🏻‍♀️</p><p>I can understand their frustration. They hoped for a live reporting dashboard that would save time, but instead, they ended up wasting even more time comparing reports to assist a solo developer with his tool…</p><p>One other French expert was passionate about Amazon PPC and tools. We had a long discussion about Amazon ads and tools. He showed me Sellerboard. A SaaS tool that helps sellers and agencies discover their profits, with detailed Amazon fees, for only $16 / month. He told me that this tool was great but he had some improvement ideas. I did not want to work on a profit calculation tool, nor to develop a “cheap” tool. My “instinct”, or actually mix-of-all-heard-things-and-experience, told me I rather stick to trying to address reporting or ads issues and charge more money so I can provide a better service.<br>We kept in touch.</p><p>The other experts were using special (vendor) accounts that work with another API and would require way more work. I decided not to invest in it.</p><h3>Pivot again?</h3><p>One and a half months after my “pivot”, I did not know what to do. I continued reaching out to more and more experts, but without success. I spent the rest of my time improving/debugging the tool, testing new landing pages, and testing an “expert public profile” approach to help experts showcase their performance. <br>I kept improving my communication skills and getting less and less awkward/uncomfortable on calls 🤓.</p><p>Talking with a lot experts I realized that most ads automation tools also provide human expertise for managing ads. <br>I once had a call with a VC from <a href="https://www.samaipata.vc/">samaipata</a>. (Yeah this sentence looks “😎” but I just wrote “working on something new” on my LinkedIn profile and got contacted by the VC, Mister Mustache).</p><p>🥸 Mister Mustache looked smart and was way cooler than what I expected when I was preparing for the “call-with-a-VC- 😎”.</p><p>He told me they invest in SaaS that not only provides tech solutions but also connects two populations together, like connecting freelancers and clients. He said these kinds of SaaS gain much more value as their user base grows. It was an interesting discussion, but I was too early stage for VC funding.</p><p>That talk and the many ones I got with experts led me to think there is something to do in connecting Amazon sellers with experts and providing them with the best tool to work together.</p><h3>Here came the vision</h3><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FKgX7on3DWfGSKGK8eI%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2FKgX7on3DWfGSKGK8eI%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2FKgX7on3DWfGSKGK8eI%2Fgiphy.gif&amp;key=d04bfffea46d4aeda930ec88cc64b87c&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="326" frameborder="0" scrolling="no"><a href="https://medium.com/media/2b70e0dedafc26b1d9b5b9707573e935/href">https://medium.com/media/2b70e0dedafc26b1d9b5b9707573e935/href</a></iframe><p>I would build a tool that connects Amazon sellers with indie experts!<br>Sellers struggle with Amazon ads, experts struggle to find clients and do reporting. I would connect them, provide the best tool so they can succeed together. The tool’s value would grow considerably, as mister mustache told me!</p><p>Yeah… Too complicated to implement. <br>Thus I did not even succeed in charging $10 for a single seller.</p><p>End of pivot.<br>End of vision. <br>End of chapter.</p><h3>💡 Targeting tech guys like me</h3><p>I kept trying things, and burning my savings, working full-time on datamz.</p><p>I was coming close to the point where I would have no choice but to find a freelance mission. I had to hurry to find the right pain point to solve and start making money !</p><p>One of the first English articles I wrote on Medium was about how to synchronize Amazon products using the Selling-Partner API. It ranked first on Google on those keywords and quickly got +3k views. I got contacted by developers struggling with this API, asking me if I knew a solution to synchronize this and this...</p><ul><li>🤔 “What if I try to solve technical issues related to people building Amazon software ?”, I asked myself, since I did not have a co-founder</li><li>😃“That sounds great… Amazon SP-API is complex, there are a lot of missing data in the official API (reviews, keyword ranking …), few competitors and it is way more easier to talk and understand developers”, I answered myself, since I did not have a co-founder.</li><li>🧐“Furthermore, I will not have to struggle hours to create uggly UI”, I conclude.</li></ul><p>I convinced myself with this quick self brainstorming 🤝</p><p>I discovered Rapid API, a marketplace where you can list your API to potential users / clients 🤩</p><p>I coded an API that does keyword ranking and review count of a product, listed in on rapid API, and got 0 users.</p><p>I edited my medium articles, talked about syncing data from Amazon, and wrote with my Rapid API solution. Still got no touch.</p><p>Wrote an article about “How to get page views and sessions” on Medium that had hundreds of views but led to no contacts.</p><p>Updated my LinkedIn profile as a “Selling-Partner API expert”… nothing.</p><p>I even reached back out to Jean-Sauveur !</p><p>I remembered that he said he also wanted to develop something related to Amazon and might hire me for that. <br>He remained silent and hidden 👀</p><p>End of pivot.</p><h3>End of cash…</h3><p>The more my financial deadline was approaching, the more I was going in multiple directions, testing all and nothing and not really knowing what I was doing at the end.</p><p>I realized it would take more time than I had imagined. So I started preparing the future freelance mission…<br>I started actively searching for a developer/lead developer position in a project that involves Amazon SP-API so I can stay in the same domain. <br>I found for the perfect one on LinkedIn! <br>A US company providing a SaaS for Amazon sellers, looking for a remote lead developer. <br>I remember I was in bed scrolling on LinkedIn. It was around 1:00 a.m. and it got so excited that I rushed out of bed to apply for the position. <br>It took me 20 minutes to apply, explaining my knowledge of the SP-API, etc.</p><p>I was pretty confident my profile would fit but… they weren’t 🥸</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Ftenor.com%2Fembed%2F24930255&amp;display_name=Tenor&amp;url=https%3A%2F%2Ftenor.com%2Ffr%2Fview%2Femotional-damage-gif-24930255&amp;image=https%3A%2F%2Fmedia.tenor.com%2Fge72hB2pobQAAAAC%2Femotional-damage.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=tenor" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/c167ce53e687effd99c2942b4ea53ea1/href">https://medium.com/media/c167ce53e687effd99c2942b4ea53ea1/href</a></iframe><p>I got contacted some weeks later by the CEO of a french startup, let&#39;s call him Flow.</p><p>Flow was “flowy”. He was very nice, cool and helpful.</p><p>When I told him about my project, he did not try to convince me that working for a great company like his is better. <br>Instead, he gave me valuable advice and shared his startup story !</p><p>In 2019 he wanted to start a SaaS with his co-founder. <br>They had already worked together on several entrepreneurial projects. <br>They spent the first 2 months brainstorming and analyzing the niche they would like to launch! They figured out that there was room for improvement in the billing solutions for craftsmen. <br>None of them had experiences in that domain, yet they started investing +200k€ to build <a href="https://obat.fr/">Obat</a>: the best alternative to existing solutions.</p><p>Months after building the tool, they started marketing it and raised fund.<br>Listening to him, I realized that there is no “right way” to build a SaaS business. Their approach was the opposite of the theoretical “lean startup” method. Yet here they were 3 years after, owning a full remote company, backed by investors, that hires +40 employees across France and Spain and making more than 140k€ MRR !</p><p>He told me google ads was “the truth” for them to see if craftsman were interested and advised me to try.</p><p>I tried Google Ads on keywords like “Amazon seller reporting”. Spent ~400€. Got 0 signup. Run out of cash. Had to find a job.</p><p>Flow was looking for a backend developer with pretty classic needs: API &amp; functional development, mostly backend, full remote.</p><p>He was aware it would be a short one, 3 to 6 months max so I can save cash and go back full-time on my project. He agreed and even told me I could reach him anytime if I need any advice.</p><p>I signed 🤝</p><h3>The Severance</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_QUQGLlaKEZdzdRssRNj-Q.jpeg" /><figcaption>Back to the Severance</figcaption></figure><p>Here I was in June 2022, back working full-time for a company 😞.</p><p>The conditions were pretty good tho.<br>Nice people, full remote, and working alone on a new API project with quite comfortable deadline. <br>As every first week in a new job, I overworked to quickly prove they have made the right choice (or maybe because of a mix of stress and ego 🤔). Then I worked “normally”. Productive enough so they would be happy and recommend me if needed, but not doing extra hours.</p><p>I slowed down on datamz that summer. <br>I just kept reaching sellers sometimes but stopped adding more features.</p><p>I realized it was the first period since 2019 that I really slow down on “working-on-a-project” mode. I managed to stop the “you-need-to-keep-working-hard-to-succeed” voice and listen more to the “oh-a-nice-burger-lets-eat-it” one 😂<br>It was a peaceful summer !</p><h4>I got into a dangerously comfortable routine.</h4><p>Everything was cool and stressless… actually too cool. <br>At that time I was watching the TV show “Severance” (great one by the way ! I recommend it !) and was feeling like I was in a kind of Severance myself 😅<br>Trying to be happy working in a position that would clearly not lead me to where I want to be.</p><p>Same routine each day, selling my time:<br>9:00 Slack message to show you are online “Hello 👋🏿”. <br>9:10 daily meeting. <br>12:30 “bon appétit” message on slack. <br>13:29 “I’m back” message. <br>18:00 “Have a good evening” message.</p><p>My routine looked exactly like that 😂:</p><p><a href="https://www.tiktok.com/@jaykimchung/video/7271012581035281710">https://www.tiktok.com/@jaykimchung/video/7271012581035281710</a></p><p>I wasn’t really learning nor taking risks and most important, I wasn’t pushing forward on my project.</p><p>I had to get back on track !</p><h3>All about meeting… and listening.</h3><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Ftenor.com%2Fembed%2F4850110&amp;display_name=Tenor&amp;url=https%3A%2F%2Ftenor.com%2Fview%2Fthe-simpsons-im-listening-listening-attentively-attentive-gif-4850110&amp;image=https%3A%2F%2Fmedia.tenor.com%2FQ3p7p9xsmp8AAAAC%2Fthe-simpsons-im-listening.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=tenor" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/0975f499f6f61a2d48d928e31b2a7090/href">https://medium.com/media/0975f499f6f61a2d48d928e31b2a7090/href</a></iframe><p>Regarding datamz, I did not know exactly which problem to solve anymore. But I was willing to connect with the right people from the Amazon ecosystem and listen to them.</p><blockquote>I remembered connecting months earlier with Mustapha, a France based Amazon seller.</blockquote><blockquote>Like almost all people I reached, I had to insist a bit and he seems annoyed to accept a video call. But like all people I had on call he finally was happy to be really listened.<br>This “quick 10mn call” ended up in a one and a half hour discussion!</blockquote><blockquote>He shared his Amazon story to me. His trouble starting directly in the US market, the way he persevered, and the main problem he faced: “Amazon ads was just ruining my profit margin!” <br>He then told me about his solution to this problem: “This is the best Amazon ads content I have seen !” he said, talking about a french trainer’s courses.<br>This french trainer focuses on Amazon Ads courses. Let’s call him Jean-Sauveur 2.<br>Jean-Sauveur 2 looks like a teacher. He is very energetic and passionate. A teacher on steroid! <br>Mustapha was passionate about JS2 courses, his methods and showed me his google sheet dashboard. <br>The process and workflow to manage his Amazon ads campaign was pure logic and very smart ! I learnt a lot during that call.<br>But the google sheet dashboard was not convenient to use. He needed to manually export report from Amazon each week.</blockquote><blockquote>Mustapha told me it would be great if he could get rid of the google sheet part and have the same dashboard directly in a tool. <br>At that time it seemed too much work investment for a single person so I reached JS2 to see if his other student would be interested.</blockquote><blockquote>He did not answered. 👀</blockquote><p>I reached out again to Jean-Sauveur 2.<br>But this time I had another approach. I just asked if I can get his thoughts on something that might his community.</p><p>He did not answered. 👀</p><p>But this time I had also learned that “no answer” doesn’t mean “no” (thanks <a href="https://www.youtube.com/watch?v=oTtBHev9cCw">Bernard Huang</a>).</p><p>I listened to Bernard’s advice and followed back every week! <br>I felt like I was harassing him but… “I was just following Youtube advice”, I would have said in court 😂</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2Fkgh6tcNzwBSQabHUYB%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2Fkgh6tcNzwBSQabHUYB%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2Fkgh6tcNzwBSQabHUYB%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="435" frameborder="0" scrolling="no"><a href="https://medium.com/media/ce5478ba7a03fc56e91e9274adf4d2e9/href">https://medium.com/media/ce5478ba7a03fc56e91e9274adf4d2e9/href</a></iframe><p>It took me a month of following up with different kind of messages to finally… get an answer !</p><p>🤩 He was ok for a quick call!</p><p>I knew I could not mess up on that one so I used all my previous call experiences and learnings (the Mom Test etc), started the call with one sentence, and just listened to him:</p><p>“I am an ex-Amazon seller, building a tool for sellers. I can build something that helps your community, I just need to get your thought their about what could help them.”</p><p>Jean-Sauveur 2, unlike Jean-Sauveur 1, behave like a real savior ! 🦸‍♂️</p><p>He got excited, told me he will help me, shared a lot of advice and immediately proposed a way to help me bootstrap !!</p><ul><li>“… and of course I will give you affiliate commission”, I propose</li><li>“No no no. I dont need any commission! <br>You need to start getting users, feedback, build a community, and I am happy to help you with that! <br>All I will ask you is to develop a free feature for my community so they can export their Amazon reports more easily. <br>But we will discuss that later! <br>In 3 weeks I will host a one week PPC challenge. One week of free Amazon Ads courses in live. <br>There are +700 sellers registered so far! <br>I will also introduce some experts to them and you could be one of them!! You will present your tool, and we will offer free beta users! The goal is for you to get a maximum of beta users! Are ou in ?”</li><li>“😱😁🤩🥳”, I thought.</li><li>“Yes sure 😀”, I said.</li><li>“Make sure your onboarding is okay, I will tell you more about how we can do to get a maximum of users”, he continued</li></ul><p>He gave me more marketing and sale advice, and discussed a plan to get the maximum of beta users. I could hear and see how excited he was about marketing and helping someone who struggle with his business.</p><p>He had almost nothing to gain but still took time to help me and give me this opportunity.</p><p>I had now the best opportunity to maybe get hundreds of beta testers! <br>To be introduced by a trusted trainer and pass the “trust” filter that is so difficult to pass at the beginning.</p><p>It was in October 2022. I had 3 weeks to make the tool ready and I still was working at the Severance.<br>I worked nights and weekends to fix some bugs and create the best onboarding experience possible before the week challenge.</p><p>6 months before, I was granted for french grant, giving 15k€ for starting the innovative project. This fund was intended to hire a work-study developer, freeing up my time for marketing ans sales.</p><p>I received that money in my company bank account around that time. Perfect timing!</p><p>It was the moment to search for a developer, nearly time to finally onboard a lot of beta users, and finally, the moment to launch the project 🚀</p><p>I was ready to welcome those hundreds of users!</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2F3oEjI1erPMTMBFmNHi%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2F3oEjI1erPMTMBFmNHi%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2F3oEjI1erPMTMBFmNHi%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="243" frameborder="0" scrolling="no"><a href="https://medium.com/media/1d823cac06ecfe2f5bc19db7781064cd/href">https://medium.com/media/1d823cac06ecfe2f5bc19db7781064cd/href</a></iframe><p>___</p><p>PS: Thank you guys for commenting my last post! Thanks for encouraging me to share my bootstrapped SaaS story. I hope it helps !<br>Writing these articles helps me improve my English and take a step back on this project. <br>But most importantly, it led me to connect with fellow entrepreneurs !<br>Next week I will be meeting an American Amazon SaaS founder (hi Ben 👋🏿) in Paris! ️🤩 The magic of sharing…</p><p>Let’s get in touch on <a href="https://www.linkedin.com/in/nassuf-mmadi/">LinkedIn</a>, share our stories, successes, mistakes, to help each other out !</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cf38dae7d1e1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The reality behind bootstrapping a SaaS • Part 1]]></title>
            <link>https://medium.com/@nassuf/growing-a-bootstrapped-saas-is-so-hard-d89f9990eb7f?source=rss-309baa13bdc3------2</link>
            <guid isPermaLink="false">https://medium.com/p/d89f9990eb7f</guid>
            <category><![CDATA[amazon-fba]]></category>
            <category><![CDATA[bootstrapping]]></category>
            <category><![CDATA[saas]]></category>
            <category><![CDATA[development]]></category>
            <category><![CDATA[solopreneur]]></category>
            <dc:creator><![CDATA[Nassuf Mmadi]]></dc:creator>
            <pubDate>Mon, 17 Jul 2023 09:47:22 GMT</pubDate>
            <atom:updated>2023-08-21T09:42:37.469Z</atom:updated>
            <content:encoded><![CDATA[<p>What you are about to read is the opposite of a “flex” article.<br>I will not tell you how I bootstrapped a $83 333 MRR SaaS in 3 months.<br>Nor will I give any advice on how to build a SaaS.</p><p>I will just share my experience, struggling on a SaaS since one and a half years. The real long path of bootstrapping a SaaS. I hope it will help you and allow me to connect with fellow SaaS developers, share experiences, and help each other out.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FwGuHmLlaTbB6gaemBF%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2FwGuHmLlaTbB6gaemBF%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2FwGuHmLlaTbB6gaemBF%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="326" frameborder="0" scrolling="no"><a href="https://medium.com/media/467556b49c6ed6680552e5cd505e365f/href">https://medium.com/media/467556b49c6ed6680552e5cd505e365f/href</a></iframe><p>I started developing this SaaS since September 2021. <br>Here are the results so far: <br>229 registered users, 375 Amazon stores, 1 400+ store’s countries, millions of orders synchronized…</p><p><strong>0 profit.</strong></p><p>Even worst. <br>Today, July 14th 2023, the company is burning €1200+ per month.</p><h3>How it started.</h3><p>In 2019, a friend (Stephen 👋) told me about selling on Amazon. He paid for courses, started selling on the marketplace and went pretty good.<br>He shared with me everything he knows about it, gave me golden advices and shared his courses so I can learn and start an Amazon business.</p><p>Few months later I was able to start selling <a href="https://www.amazon.fr/dp/B07T924DV6/">yoga mats on Amazon</a>, thanks to him !<br>Getting the first sales is an amazing feeling ! 🤑<br>You feel like you finally succeeded in turning an idea to an actual product that people pay for. <br>Then I got my first real good reviews ! A few months earlier I didn’t knew anything about selling on Amazon, even less about yoga, and here I was, selling dozens of premium yoga mats to happy customers.</p><p>I felt like the yogi version of Steve Jobs 😅</p><p>I was feeding my ego with a well-growing revenue. I used to talk to my friends about how many mats I sold each day and how great of an opportunity Amazon is. I felt like a visionary entrepreneur who knows things…</p><p>Stephen advised me to use ShopKeeper, a SaaS tool that tracks all Amazon costs so you can know your real profit.</p><p>And it was at this moment that I knew… 💥</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qUY6g_lHwZdEzHm5nC2G7w.jpeg" /><figcaption>Discovering my actual profit on ShopKeeper</figcaption></figure><h3>10k€ loss</h3><p>Selling on Amazon is all about ranking your product for keywords.<br>Those keywords must be relevant enough for people to buy your product and, of course, have enough search volume.</p><p>The yoga mats I was selling were built for experienced yogis. I noticed in competitors’ product reviews that a lot of yogis complained about slippage. Yoga postures require a mat with good grip, so I opted for premium natural rubber mat material with an ultra-grippy layer on top.</p><p>The main keywords I was targeting (in FR) were:<br>- Yoga mat<br>- Natural rubber yoga mat<br>- Anti-slip yoga mat<br>- Ultra-grip yoga mat</p><p>Well, to rank for those keywords, you mainly have one choice: Amazon Ads.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xh1JS_2qzCWEgEYX7AokCw.png" /><figcaption>What a sponsored product looks like</figcaption></figure><p>Amazon Ads’s concept is quite simple:</p><ol><li>You target a keyword and set a bid, let say €1.50.</li><li>Each time a visitor clicks on your “sponsored product” you get charged €1.50.</li><li>Your organic ranking on that keyword improves each time a visitor buys your product after clicking on your ad.</li><li>You are supposed to lower your bids once your organic rank is good.</li><li>You get more and more organic sales, more and more profit and eventually become a fackin-multi-millionaire <strong>🤑</strong></li></ol><p>Quite simple in theory.</p><p>In practice it is way more complicated.<br>You quickly spend a lot on ads to figure out which keywords are relevant. Sometimes it works, sometimes not, and you dont know why.<br>You end up paying for Ads courses and learn about strategies appear efficient but are too complicated to apply on a daily basis…</p><p>Fast forward, I made a lot of mistakes, lost a lot on ads and failed.<br>⚰️ 2019–2021. 1400 mats sold. €85k+ revenue. <strong>€10k loss</strong> !</p><p>I stopped selling on Amazon and, as I am a developer, started a SaaS to help Amazon sellers with issues I faced.</p><h3>🚀 September 2021, let’s go Harissa Man !</h3><p>Developing an ads automation tool is quite long and complicated. <br>So to start quickly and “lean”, I focused on another small issue/niche related to Amazon sellers.</p><p>While I was selling on Amazon, one technique to boost your product ranking was using “Supreme URL”.<br>Basically, a service called pixelfy.me let you generate small URLs that manipulated Amazon’s URL parameters to make them believe that the visitor had typed a certain keyword. If the visitor purchased your product, it was supposed to improve your ranking. <br>Kinda borderline, but you gotta do what you gotta do right ? 🏴‍☠️</p><p>The problem with Pixelfy was that it wasn’t possible to track your keyword rank and actually see if the Supreme URL was effective.</p><p>💡 I had my MVP idea !</p><ol><li>I would develop a Pixelfy.me copycat, with a smoother UX, including a Supreme URL + keyword rank tracker. I estimate it would take around 3 weeks of development.</li><li>Then, I would talk about it on the 500+ Amazon sellers Discord group I was a part of, sending messages to sellers one by one with a ready-to-use product.</li><li>I would also mention it to the course trainer (who runs the YouTube channel “Money TV”), let’s call him French-Money-Fox.</li><li>Eventually French-Money-Fox would find the service very helpful and share it with the community in exchange for affiliation.</li><li>The community would be excited to try it for free, love it and pay for it. I would get dozens, then hundreds of customers and then develop the ads automation feature, and eventually become fackin-multi-millionaire. 🤗</li></ol><p>It is more fun to work with people, so I talked about it with my friend HM, Harissa Man. I decided to take this opportunity to try a new framework, Laravel, which is loved by HM.</p><p>HM was ok 🤝</p><p>We started building the MVP.<br>He developed all the clean foundations and assisted me make the transition from Symfony to Laravel.</p><p>A month later, the MVP was ready and so were we.</p><p>I shared it with the Discord community, sending personalized messages to sellers one by one.<br>I also sent a (way too long) message to French-Money-Fox.<br>Nothing happened. <br>No trust. No answer. No user. Even for free.</p><p>I changed my approach and start engaging more with sellers before introducing them to the tool.</p><p>✅ We got our first beta testers.<br>❌ We got that answer from French-Money-Fox: “Looks cool, but I don’t have time”<br>💥 Harissa-Man grew tired of working on a tool that didn’t provide fast enough returns and decided to quit.</p><p>He left solid technical foundations and continued to offer advice and provide code reviews for months. 🙏</p><p>🦸 He wished me luck and ran fast across the street with a harissa sandwich.</p><h3>The Mom feedbacks period</h3><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FTkfbLsVsFiiEnQKHus%2Fvideo&amp;display_name=GIPHY&amp;url=https%3A%2F%2Fgiphy.com%2Fclips%2Fguardian-the-guardian-TkfbLsVsFiiEnQKHus&amp;image=https%3A%2F%2Fmedia1.giphy.com%2Fmedia%2FTkfbLsVsFiiEnQKHus%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="480" height="270" frameborder="0" scrolling="no"><a href="https://medium.com/media/d1fa57e0b130441ed27362ecdff70999/href">https://medium.com/media/d1fa57e0b130441ed27362ecdff70999/href</a></iframe><p>Reality is often far away from theory and logic.</p><p>In theory, sellers who are already using Pixelfy.me would be willing to try a better solution for free.<br>In reality, they tend to only try solutions they trust, and won’t trust a came-out-of-nowhere ex-seller/developer proposing a free solution.</p><p>Moreover, selling on Amazon is all about finding good niches, and the level of trust and value must be strong enough for sellers to disclose their niches to a potential competitor. 🥷</p><p>Being too excited with the genuinely better solution we built, I underestimated how important trust would be.<br>I also underestimated how challenging it would be to convince “trust-owners,” such as Amazon trainers, to try and promote the tool in exchange for affiliation.</p><p>Here began the “travail de fourmis” (ant work) phase. <br>I spent days across Discord servers and Facebook groups, sending hundreds of personalized messages.<br>Most sellers never responded, and only a few registered but never really used the tool.</p><p>I realized that I was clearly doing something wrong in my approach, but I couldn’t figure out what it was.</p><p>I continued reaching out to more Amazon trainers and YouTubers. One well-known figure (<a href="https://www.youtube.com/@DanRodgers">Dan Rodgers</a>) was kind enough to register but never used the tool. I believe the tool’s functionality was simply not strong enough, and the main feature of tricking the Amazon algorithm was seen as too risky or questionable.</p><p>I finally got in touch with a french trainer who has a community of over 200 students!<br>Let’s call him Jean-Sauveur.</p><p>Jean-Sauveur is not on Youtube. <br>He doesn’t over promise about succeeding on Amazon. He makes all his acquisition from Google SEO &amp; Google Ads. He seems well-organized, and his courses appear to be serious and reputable. One of my friends actually paid for his Amazon FBA training and was satisfied with the experience.<br>- “His course is very clear, and he takes the time to have a call with each of his students before they start the training.”, she told me.</p><p>I had a call with Jean-Sauveur.</p><p>He didn’t show much enthusiasm about “Supreme URL” tricky url feature. <br>So I told him about the next feature I planed to add: <br>✨ A dashboard to have a clear and simple overview of key metrics. ✨<br>He became excited about it !</p><p>- “Do you think it would help your students?”, I asked<br>- “Yes, for sure”, he said<br>- “If I develop it, will you promote the tool to them with affiliate links?”<br>- “Of course, once the dashboard is ready!”</p><p>I was so excited ! The first Amazon trainer to validate the tool I was building ! A first door opened to, hopefully, tens of sellers. 🤩</p><p>I stopped sending messages to individual sellers and dived into the Amazon “Selling-Partner API” to connect with Amazon sellers’ data and build the first dashboard.</p><p>At this moment, I discovered what a mess this API was. <br>Getting access to it took me an entire week! <br>Synchronizing the seller’s products took me 3 weeks!</p><p>It was so unreal for such a big company like Amazon that I wrote <a href="https://medium.com/@nassuf/how-to-get-a-sellers-listings-using-amazon-sp-api-475771d3754c">an article to document it</a>. The article quickly ranked well on google, got hundreds then thousands of views and some connection with fellow developers struggling like me. It later led me to try focussing on this audience, but I will talk about that in another article.</p><h4>Time to launch !</h4><p>It took me a month to build the dashboard. I ended up having the exact dashboard that I wished I had when I was selling on Amazon. 🤩</p><p>I immediately reached out to Jean-Sauveur !</p><blockquote>“Hey Jean-Sauveur ! <br>Hope you are well. The dashboard is now ready ! Sellers can see all of their key metrics in a single view ! All countries and all ads combined ! They can even see their organic sales, which is important yet not available on the official Amazon seller dashboard (Amazon Seller Central).”, I sent.</blockquote><p>He never answered.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2Fjc2PkKKr3clTBekMzn%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2Fjc2PkKKr3clTBekMzn%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2Fjc2PkKKr3clTBekMzn%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="244" frameborder="0" scrolling="no"><a href="https://medium.com/media/9a875b4064ff622e90f6aa9117360bd6/href">https://medium.com/media/9a875b4064ff622e90f6aa9117360bd6/href</a></iframe><p>LinkedIn, email, phone, I tried them all. <br>Jean-Sauveur disappeared. <br>He was not the savior I expected.</p><p>Started one of those painful periods in entrepreneurship: <strong>doubts</strong>.</p><p>What went wrong ? What mistake did I make ? Am I in the right way ? Am I even solving a real issue ? Why some sellers tell it’s a good idea but few registered ? Am I that bad at “selling” ? Maybe Harissa-Man left because he knew we wont make it. Why Jean-Sauveur disappeared ? Did he lied ? Why ? Is it because I am black ? Am I black ? 😅</p><p>I obviously had to change something. I had to learn how to talk to sellers and trainer to know their real issues.<br>I searched for books about that on Youtube and discovered a <a href="https://www.momtestbook.com/">gold nugget</a>…</p><h3>The Mom Test</h3><p>Do you know this moment when you are stuck for so long in a video game and you get tips that unlocks everything?<br>That is exactly how I felt with this book!</p><p><a href="https://www.momtestbook.com/">The Mom Test</a> teaches you “How to talk to customers &amp; learn if your business is a good idea when everyone is lying to you.” (especially Jean-Sauveur) !</p><p>I learned that I asked questions in the worst possible way, which left sellers and trainers with almost no choice but to answer “yes, that is a good idea.” I learned that there is a world between saying “yes, I will use it” and actually using it. I learned how to conduct user interviews, discover real issues, and identify when I was “demonstrating” or approaching things incorrectly.</p><p>I strongly recommend this book to anyone willing to launch a product. It was the best book I have read in 2023 so far (and actually the only one 😅).</p><p>I also discovered one of the best talks I have seen about getting your first customers. <br>No theory, a highly practical talk that shares insightful experiences. I recommend watching it.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FoTtBHev9cCw%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DoTtBHev9cCw&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FoTtBHev9cCw%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/d582c0e12d0b70f52ee21939335f96b4/href">https://medium.com/media/d582c0e12d0b70f52ee21939335f96b4/href</a></iframe><h3>Pivot ?</h3><p>Finding and convincing sellers to have a call was too difficult.<br>I failed to have calls with trainers to listen to their student issues.<br>I even started doubting the Amazon ads automation struggle. Yes I struggled with it as a seller, but I never paid for an automation software myself.<br><br>Among the people I met, there was a freelancer who worked with big companies selling on Amazon. Thanks to the mom test method, I discovered that his main issue was extracting data from Amazon to create his weekly report.</p><p>I contacted most of the French Amazon freelancers, added hundreds of Amazon experts on LinkedIn with an automation tool and sent them personalized message to have a talk.</p><p>I eventually managed to talk to over 100 experts. They almost all have two main issues:</p><ol><li>Struggle to find new customers (sellers)</li><li>Long time spend on extracting data for weekly / month reports</li></ol><p>I though: What if I focus on Amazon experts instead of sellers ?</p><ul><li>They are way more easier to target (LinkedIn).</li><li>I identified a first issue that can be easily solved (online reports)</li><li>They might have a larger budget than most sellers.</li></ul><p>It was time to pivot !</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FnL76f9syWJBoA%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2FnL76f9syWJBoA%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2FnL76f9syWJBoA%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="261" frameborder="0" scrolling="no"><a href="https://medium.com/media/dcb46d41f9ce2d4fc9513e8020755b4b/href">https://medium.com/media/dcb46d41f9ce2d4fc9513e8020755b4b/href</a></iframe><p>Let me know if it’s worth writing what happened next.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d89f9990eb7f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[ChatGPT is a bullshitter. (shit-ton-money-journey-part-2)]]></title>
            <link>https://medium.com/@nassuf/chatgpt-is-a-bullshitter-shit-ton-money-journey-part-2-9562de3e5587?source=rss-309baa13bdc3------2</link>
            <guid isPermaLink="false">https://medium.com/p/9562de3e5587</guid>
            <category><![CDATA[money]]></category>
            <category><![CDATA[openai]]></category>
            <category><![CDATA[developer]]></category>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[saas]]></category>
            <dc:creator><![CDATA[Nassuf Mmadi]]></dc:creator>
            <pubDate>Sat, 21 Jan 2023 14:47:34 GMT</pubDate>
            <atom:updated>2023-01-21T14:47:34.250Z</atom:updated>
            <content:encoded><![CDATA[<p>In the <a href="https://medium.com/@nassuf/making-a-shit-ton-of-money-with-chatgpt-experiment-1-52c60ec8037f">previous article</a>, I shared with you guys how I came out with a “ChatGPT idea”.<br>The idea to create an automated blog with articles that recommends books based on Twitter trends… With affiliate links 🤑<br>A blog that can generate a shit-ton of dollars. Or not 😅</p><p>To be honest I did not expect that many interactions ! So thank you guys for encouraging me to keep writing !</p><p>Here is what happened next.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2Fw97X9z4X3O4hi%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2Fw97X9z4X3O4hi%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2Fw97X9z4X3O4hi%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="326" frameborder="0" scrolling="no"><a href="https://medium.com/media/531eaa5f96b892d48d9d91c8f029cc5e/href">https://medium.com/media/531eaa5f96b892d48d9d91c8f029cc5e/href</a></iframe><h3>ChatGPT, the bullshitter.</h3><p>As you understood, the goal is to have fun. <br>To try to build things with great technologies, and share the story with you.<br>Hopefully it will make money but first it need to be something cool to use.<br>Something that really helps, with <strong>trustful</strong> content.</p><p>So, we have 3 steps:</p><p>✅ We get Twitter trends via its API<br>✅ We generate an article via OpenAI API with a detailed prompt<br>✅ We scrape Amazon to get related books</p><p>From a technical point of view it perfectly worked !<br>I used <a href="https://github.com/timlrx/tailwind-nextjs-starter-blog">timlrx/tailwind-nextjs-starter-blog</a> for the blog as we can embed markdown articles directly in it.<br>The soft (Laravel command) just have to generate markdown formatted articles, put them in the right folder and that’s it !</p><p>Technically I could have created a fully automated blog.</p><p>But…</p><p>⚠️ OpenAI’s recommended books topics were sometime irrelevant<br>⚠️ GPT was <strong>often totally bullshitting</strong> when summarizing books 😅</p><h4>Here is an example</h4><p>The current top trend in San Francisco while I am writing this article (Jan 20th) is “Warriors”, the basketball team.</p><p>Here are the 5 selected tweets to build the prompt.</p><p>Those tweets will give GPT enough context to write a relevant article.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Xxhug41z55utetY6JiIwng.png" /><figcaption>1. Get Twitter trends, select 5 tweets that gives context</figcaption></figure><p>Based on the prompt, it has generated an article and added 3 topics to be searched on Amazon:</p><ol><li><strong><em>__Watching the Warriors play__</em></strong> that led to the book <a href="https://www.amazon.com/dp/0578660636/">Mental Toughness For Young Athletes</a> 👍</li><li><strong><em>__Learning about basketball strategy__</em></strong> leading to <a href="https://www.amazon.com/dp/0989746607/">The A-Z Basketball Book</a> 👍</li><li><strong><em>__Reading about the history of the Warriors__ </em></strong>for which the first book result on Amazon is… <a href="https://www.amazon.com/dp/B08Y4K6SNT">The Best is Yet to Come: The Memoir of a Peaceful Warrior</a> 💩</li></ol><p>In this example the last book is totally irrelevant.</p><p>For each book, the soft calls GPT API to get a friendly summary. <br>The last book’s summary is even more annoying.</p><p>Here is the third book’s description on Amazon:</p><blockquote>🔫 After 41 years in the U.S. Army and the private sector, as a “Warrior”, Colonel (Retired) Rick Kiernan has witnessed American History “first-hand” from the <strong>Cold War to Vietnam, to the Gulf War, and finally, the Global War on Terrorism</strong>. His experiences involved sharing sacrifices with a “band of brothers”. In this memoir, which has been a “labor of love”, <strong>he describes his adventures during his years in the service</strong>, a journey which has taken him to live in ten states and six countries. He also includes “lessons learned” in an effort to “pass the torch” of service to others who may serve their country and the values that we hold so dear but always have to defend from all enemies foreign and domestic.</blockquote><p>Now here is the GPT sum up, from the prompt <em>‘summarize the book [BOOK_TITLE] to make people want to read it.’</em></p><blockquote>☮<em>️</em> The Best is Yet to Come: The Memoir of a Peaceful Warrior is an inspiring story of one man’s journey to discover inner peace and fulfillment. Through his life experiences, <strong>he learns to embrace the power of meditation, mindfulness, and self-love.</strong> He shares his wisdom and insights on how to live a life of joy and purpose, and how to create a better world for all. This book is an uplifting and inspiring read that will motivate readers to find their own inner peace and strength.</blockquote><p>It’s false, it far from the book’s story and has nothing to do with the trend.</p><p>Untrustable GPT.</p><h3>Why is GPT lying ?</h3><p>While looking to understand how GPT works and why it produces fake answers that looks real I discovered a great Youtube video ! (🇫🇷 enable subtitles)</p><p>I learnt that:</p><blockquote>“ChatGPT does not answer to questions. It just generates a credible sequel to your prompt.”</blockquote><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FR2fjRbc9Sa0%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DR2fjRbc9Sa0&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FR2fjRbc9Sa0%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/7ea70a4e2760c600d6176ae6f1a18da2/href">https://medium.com/media/7ea70a4e2760c600d6176ae6f1a18da2/href</a></iframe><p>They basically explain that ChatGPT is not meant to answer right to your questions. It “completes” your prompt with the most likely sequel.<br>Like your phone auto-completion, but trained with huge data.</p><p>By the way, write a word and keep choosing the middle word (auto-complete) in your phone, you will discover your own “mini-GPT” 😀</p><p>Here is mine, starting with “I”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TmXJJJp9e9Cqmw7Q_15ZRw.jpeg" /><figcaption>Share yours in comment 😅</figcaption></figure><p>That explains why it wrote a credible yet totally wrong summary for the last book.</p><h3>So, how to make it tell the truth ?</h3><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2F3o6UBdd79rm7vxtOSs%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2F3o6UBdd79rm7vxtOSs%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2F3o6UBdd79rm7vxtOSs%2F200.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="220" frameborder="0" scrolling="no"><a href="https://medium.com/media/c882ae04a78d42ebf6cf86a543ddbb74/href">https://medium.com/media/c882ae04a78d42ebf6cf86a543ddbb74/href</a></iframe><p>We have been blown away by its capacity. Now we have to understand its limits and the way it works so we can get the best out of it.</p><p>We now know that it is completing, not answering.<br>We know that it can help us a lot when it comes to writing content but not totally replace us, nor be fully trusted.</p><p>We must therefore have two important roles: <strong>Guide and Guard</strong>.</p><p>The more we <strong>guide</strong> GPT, with precise and contextual information, the more relevant the content will be.</p><p>That said, I think we will always have to be the <strong>guard</strong>. <br>We <strong>must</strong> check the result before spreading it.</p><p>By the way, <a href="https://www.youtube.com/@Science4allOrg/about">Dr. Lê Nguyên Hoang</a>, computer science researcher, who talks in the video I shared above, said:</p><blockquote>We can imagine that the next generations of ChatBot will not only seek to predict what a human would write, but increasingly to predict what human editors prefer (among the things it predicts that a human would write). This is already quite clearly the case with GPT-3.5, which is trained to avoid certain topics…</blockquote><h3>📚 Back to the blog</h3><p>Guide and guard…<br>To ensure that the articles are interesting and without irrelevant books recommendations, I decided not to auto-publish.</p><p>Instead, I put several checks and validations during the process.</p><p>Here is the current one:</p><ol><li>The command shows each trend + 5 tweets and let me choose if we should write an article about it.</li><li>It send an OpenAI prompt to get an article and 3 related topics.</li><li>It then displays the topics. I can validate or edit each topic before it’s being searched on Amazon.</li><li>The command scrapes Amazon’s result page searching for the topic in the book section. It gets book’s title, images and urls.</li><li>Instead of auto-selecting the first book, it shows me the first 10 organic ranked (avoiding sponsored ones) and let me choose which one should be included.</li><li>I personally check on Amazon and select the right one.</li><li>It calls OpenAI again with the book title, asking a sum up linked to the trend.</li><li>Shows me the result that I once again can edit and validate.</li><li>Generate the article, at the right place to be ready for publication.</li><li>I just have to recheck the article and publish it.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*P73s3xjOJfsGNrZrVibrjg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fXNhPfYo_EWAmSM4WEDHTw.png" /></figure><p>Doing so, it takes around a minute to create an article, on a trendy topic, with 3 books recommendations, fully checked by a human.</p><p>Good enough ✅</p><h4>Does it work ?</h4><p>For the moment, absolutely not 😅</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*C_o0sG_CkdzyvwKf2TL3zA.png" /><figcaption>Google Analytics caption.</figcaption></figure><p>The last 14 days I published 36 articles, including up to 3 books each.<br>Today, only 7 of those articles are indexed on Google.</p><p>That is the reality behind what seems to be a great idea vs real world execution. The reason why <a href="https://medium.com/@nassuf/why-you-will-not-make-money-with-chatgpt-written-by-chatgpt-81acea6eed2c">most of us will not make money with ChatGPT</a>.</p><p>In order for the blog to work, I have to keep pushing. Find ways to bring the first users.</p><p>I could develop a Twitter bot that uses GPT to respond to related tweets and spread the articles. <br>I could find other sources of trends.<br>You also certainly have a lot of ideas that worth sharing…</p><p>But as it is a side experimental project, I can’t spend too much time on it.</p><p><strong><em>Any developer wants to join me in this ? </em></strong><em><br>(There is a shit-ton-of-money to make </em>😄<em>)</em></p><h3>Shares</h3><p>I wanted to share the blog link with you guys but as I just started writing english articles today instead of french, I want to let it run one more week to see how it ranks, without medium’s external visits. <br>I will share it in a few days</p><p>Meanwhile, a friend sent me a very funny project mixing midjourney / OpenAI GPT and Tiktok.</p><p>A <a href="https://www.tiktok.com/@genepix.ai">Tiktok account with short stories</a> written and drawn by AI. <br>People submits stories in comment. Each day they select one and post the result the day after.</p><p>It is only in French tho. You could do the same in English.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*SOUIvAVoyHMccSBzSAqgIg.png" /><figcaption>“Story of a goat becoming a drug dealer” 😂</figcaption></figure><p>Share the best projects, videos and books you know !</p><p>I will keep you guys updated on this STOM-journey 😄</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9562de3e5587" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Making a shit-ton of money with ChatGPT. Experiment 1.]]></title>
            <link>https://medium.com/@nassuf/making-a-shit-ton-of-money-with-chatgpt-experiment-1-52c60ec8037f?source=rss-309baa13bdc3------2</link>
            <guid isPermaLink="false">https://medium.com/p/52c60ec8037f</guid>
            <category><![CDATA[developer]]></category>
            <category><![CDATA[openai]]></category>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[money]]></category>
            <category><![CDATA[saas]]></category>
            <dc:creator><![CDATA[Nassuf Mmadi]]></dc:creator>
            <pubDate>Tue, 10 Jan 2023 00:38:55 GMT</pubDate>
            <atom:updated>2023-11-17T16:45:59.466Z</atom:updated>
            <content:encoded><![CDATA[<iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2Fh0MTqLyvgG0Ss%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fgiphy.com%2Fgifs%2Fmorning-tea-wednesday-h0MTqLyvgG0Ss&amp;image=https%3A%2F%2Fmedia1.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExeDBzOWdnNDN5NWZiMjA1bGZudzN1YWdzenQ2MjNuNmFoajNuMnZpNyZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw%2Fh0MTqLyvgG0Ss%2Fgiphy_s.gif&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="331" frameborder="0" scrolling="no"><a href="https://medium.com/media/079b911ebf51ed2b45b7a6f8ef548697/href">https://medium.com/media/079b911ebf51ed2b45b7a6f8ef548697/href</a></iframe><p>As many of you I have been blown away by ChatGPT. <br>I spent too much time playing with it. Generating articles and code. Analyzing amazon products. Using it as an assistant developer. And discussing during hours with friends trying things with it.</p><p>I got excited about trying OpenAI API, just for fun.<br>But as I am working on <a href="https://datamz.me/">an early stage SaaS</a> that doesn’t make revenue yet, with very short deadlines, it wasn’t reasonable to “lose” time on it.</p><p>Unless… I find <strong>a way to make money while having fun.<br></strong>Unless doing this experiment would led me produce code that can be used in our project.</p><p>Thus I made a deal with “reasonable me” 🤝</p><ol><li>I must do it at night, he said. After 10pm, when I am done doing what needs to be done on datamz.</li><li>It must be done in less than 3 evenings.</li><li>It must serve a concrete immediate need on datamz.</li><li>I must include things that can be re-used in short term on datamz.</li><li>“Well… at least try to make money with it… and create content about it”, he conclude.</li></ol><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2F3ohzdS6pp7KWLCO0r6%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2F3ohzdS6pp7KWLCO0r6%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2F3ohzdS6pp7KWLCO0r6%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="181" frameborder="0" scrolling="no"><a href="https://medium.com/media/8f5b9f5f443f0acab7f45c40e69d80b2/href">https://medium.com/media/8f5b9f5f443f0acab7f45c40e69d80b2/href</a></iframe><p>So here I am, Wednesday 4th evening, excited about trying new things 🤩<br>A new POC that might make <strong>shit-ton of money</strong>… or not.</p><h3>The idea</h3><p>A friend of mine runs <a href="https://www.julsa.fr/">a tech / running blog</a> where he share his passions. He tried some articles written with ChatGPT that performed very well.<br>Few months ago he tried an <a href="https://la-semelle.com/">affiliation website</a> but it took too many time to write articles. So he stopped. And I think he is less motivated writing about shoes than running and tech. Understandable.</p><p><em>💡 What if I created an automated affiliate blog ?</em><br>Affiliation can be a way to make that dollars (deal point 5. ✅) and I need to build datamz’s blog next week so I could re-use the template (3. ✅).</p><p>But what about ? What a blog could personally help me on ? And what would I buy from an article recommendation ?</p><p>… Books !</p><p>I really want to read more ! Sometimes I search on Youtube for books recommendations… I always find the same. It would be great if I can get some books recommendations on various topics. I dont want to be limited on “business, entrepreneur, tech” books.</p><p>How about getting recommendations based on hot news / trends ?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*HiomcABDI_2apsIUij9mQw.png" /><figcaption>One of the things ChatGPT can’t do for now.</figcaption></figure><p>And here are our POC idea:<br>- An <strong>automated blog</strong> that will <strong>get Twitter trends</strong>, <strong>write articles</strong> and <strong>recommend 3 books</strong> to read about it. With affiliate links, of course.</p><p>The OpenAI API client code could be re-used on datamz to bring intelligence to analytics and help Amazon sellers get great stuff fast (4. ✅). It seem like a lot of things to develop but I already have some code that scrape Amazon and I have a secret assistant that will help me write the Twitter / OpenAI API implementation (1.2 ✅). You guess it right, ChatGPT itself !</p><h3>Let’s goo !!</h3><p>The app can be divided into 3 parts:</p><ol><li>Twitter stuff. Get the trends from the API and everything we need to build an understandable prompt for OpenAI.</li><li>OpenAI API implementation. To send the prompt and get the article content with the 3 books recommendation.</li><li>Connection to Amazon to get the books images and affiliate url.</li></ol><p>To have more fun and go faster I talked about this idea to my friend Yassine, a fellow guru developer who preached me Laravel until I joined the cult 😅</p><blockquote>“Join-me on this Yassine ! We will make shit-ton of money !!!” I argued.<br>“No. You should eat harissa.” he politely declined.</blockquote><p>ChatGPT was more convinced.</p><h4>1. Getting Twitter trends</h4><p>It is amazing how fast you can develop basic stuff using ChatGPT.<br>Instead of reading the TwitterAPI to understand how it works, I just asked to ChatGPT to write the code and understood way quicker.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZQRB7-O7qEW6ZrUBjeGsCA.png" /><figcaption>Ready to use TwitterAPI implementation that get Trends and 5 tweet from each</figcaption></figure><p>I just had to register to TwitterAPI, apply to the right one (short form filled within a minute) edit the API credentials and it was ready !</p><h4>2. Generating the article with OpenAI</h4><p>Same with OpenAI API implementation. I just asked ChatGPT to write the PHP implementation to ask OpenAI to write and article and get the closest result as ChatGPT would return.<br>You might know that OpenAI API actually aren’t as efficient as ChatGPT. I read that ChatGPT uses a model that is not available yet on OpenAI APIs.</p><p>Yet it worked pretty well !</p><p>In less than 2 hours I had an app that gets Twitter trend, extract 5 long enough tweets, build a prompt and send it to OpenAI to write an article about it !</p><p>I even specify in the prompt that the books must be between double underscores so I can parse them to search them on Amazon later.</p><p>Here is the <em>massive</em> prompt (in French):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JzdYbOAkOxwzd3G88o2Stg.png" /><figcaption>Specified to write in in Markdown so it is ready for a blog</figcaption></figure><p>After many tries, the recommended articles were often not relevant. So instead of asking ChatGPT to recommend books, I ask to write 3 topics about the article. We will then use those topics to search on the book section of Amazon and extract books that might be interesting.</p><h4>3. Getting the right books name, image and affiliate url.</h4><p>To get and affiliation tag you just have to register for free on <a href="https://affiliate-program.amazon.com/">https://affiliate-program.amazon.com/</a></p><p>That’s what I did to got my “tag-id” that I will have to add after each book url.</p><p>Once OpenAPI returns the article content, we get each topics <em>preg_match_all(‘/__(.*?)__/’, $content, $matches);</em></p><p>For each one we just have to search on Amazon book section and get the first organic ranked book.<br>I would ask to ChatGPT for that but I already had the base-code that scrap Amazon result page.</p><p>We then remplace the topic but the book image, title and link and voilà !</p><p>That is how I built the “write-article-based-on-trends-that-recommend-books” code base in less 4 than hours.</p><p><strong>And this is what happened next: </strong><br><a href="https://medium.com/@nassuf/chatgpt-is-a-bullshitter-shit-ton-money-journey-part-2-9562de3e5587">https://medium.com/@nassuf/chatgpt-is-a-bullshitter-shit-ton-money-journey-part-2-9562de3e5587</a></p><h3>What I learned from there</h3><p>I don’t know if AI will replace developers. It might. But I know that it can be a great tool to be more efficient on value-less development.<br>I think our value should not be into implementing APIs or scrapping data. The more we can delegate those tasks, the more we can focus on trying to be more efficient. Thinking about creative solutions. Finding the easiest ways to solve complex business problems.</p><p>I think this is the starting of really interesting things. People who embrace those technologies and find ways to make it serve there solutions will take most advantage out of it and might make… shit-ton-of-money.</p><h3>Next ?</h3><p>Let me know if you are interesting in the next episode. Spoiler: The blog is online since 3 days, with already 25 articles and more than 70 books recommandation 😃</p><p>Meanwhile, let me share to you two great videos / podcast I have listened / viewed about AI and business recently:</p><p><em>Some ideas about what to build with ChatGPT:</em></p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FWZ3A2-gU2gM%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DWZ3A2-gU2gM&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FWZ3A2-gU2gM%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/8b615c018083028784d445de9099ff99/href">https://medium.com/media/8b615c018083028784d445de9099ff99/href</a></iframe><p><em>And a very interesting talk with Martin Cloake, the CEO of a SaaS that uses AI to help their client make decisions:</em></p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FAwRieeZBcCk%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DAwRieeZBcCk&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FAwRieeZBcCk%2Fhqdefault.jpg&amp;key=d04bfffea46d4aeda930ec88cc64b87c&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/5d6c7b0df6548470f731c95fa6201955/href">https://medium.com/media/5d6c7b0df6548470f731c95fa6201955/href</a></iframe><p>Share yours :)</p><h4>Edit</h4><p><a href="https://medium.com/p/f724eca0b1e3">Here is what happened next :)</a></p><p><a href="https://medium.com/p/f724eca0b1e3">https://medium.com/p/f724eca0b1e3</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=52c60ec8037f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Problem with Amazon PPC Automation.]]></title>
            <link>https://medium.com/@nassuf/the-problem-with-amazon-ppc-automation-a8d2e2b2d9c3?source=rss-309baa13bdc3------2</link>
            <guid isPermaLink="false">https://medium.com/p/a8d2e2b2d9c3</guid>
            <category><![CDATA[amazon-fba]]></category>
            <category><![CDATA[ppc-automation]]></category>
            <category><![CDATA[amazon-ads]]></category>
            <category><![CDATA[amazon-ppc]]></category>
            <category><![CDATA[ppc]]></category>
            <dc:creator><![CDATA[Nassuf Mmadi]]></dc:creator>
            <pubDate>Sat, 24 Dec 2022 01:13:48 GMT</pubDate>
            <atom:updated>2022-12-24T01:13:48.170Z</atom:updated>
            <content:encoded><![CDATA[<p>If you’re selling on Amazon, you’re probably aware that PPC (pay-per-click) advertising is a must use to be competitive, sell and rank your products.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JTH8V4ohEMCaObslAmCreQ.png" /><figcaption>From “Nanakia” Youtube live, by Cyril Bentz</figcaption></figure><p>In 2020, only 4 of the 16 first products shown in the Amazon result page were sponsored product, the rest was organic. <br>In 2021, the first half of the page was sponsored.<br>Since 2022 only 4 products are organic in the first result page !<br>So let’s be clear: <strong>You have to use Amazon PPC </strong>and be better than your competition on that !</p><p>However, while PPC can be a powerful tool for boosting the visibility of your products, it’s important to approach it with caution and ensure that your campaigns are set up and managed correctly in order to be profitable.</p><h4>A mistake I saw a lot…</h4><p>One common mistake that sellers make when it comes to PPC is running their campaigns continuously without taking a strategic approach. <br>Your PPC <strong>should help to improve your organic ranking</strong> on Amazon. Your sales on a specific keyword should pass from ads to organic over time, while your ranking improves. This way you will be able to reduce your spend on the keyword and still selling once your organic ranking is good, resulting in more profits !</p><p>It is very important to focus on profitability and ensure that your PPC campaigns are generating a <strong>positive return on investment.</strong></p><h4>The process</h4><p>One of the best way for that, taught by <a href="https://www.youtube.com/@AmazonSecretTuto">Cyril Bentz</a> is to</p><ul><li>Target a specific keyword.</li><li>Follows how you perform on it by tracking your click through rate and conversion rate.</li><li>Make sure your conversion rate is above your store average conversion rate.</li><li>Follow the organic rank of your product on this keyword while you are making sales.</li><li>If investing on this keyword does not make your organic rank improve: stop investing on it. <br>If it does, reduce your bid while you get more and more organic sales via this keywords.</li></ul><p>Cyril has taught this summarized process to more than 1000 students and guess what ? It works ! You get to know quick enough if your product is not good on a keyword. You know where to invest, when to reduce etc.<br>The only problem is… it is painful to manage !</p><p>As you might know, doing all of it manually on multiple keywords, sometime multiple products can be a real pain in the *ss.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2F7T33BLlB7NQrjozoRB%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2F7T33BLlB7NQrjozoRB%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2F7T33BLlB7NQrjozoRB%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="239" frameborder="0" scrolling="no"><a href="https://medium.com/media/fed1b0add7e57a5ed2965e38f0c4fb72/href">https://medium.com/media/fed1b0add7e57a5ed2965e38f0c4fb72/href</a></iframe><h4>Then comes PPC automation software !</h4><p>So, how do most PPC automation software tools work today? Generally, they use algorithms to determine the most performing keywords and adjust bids accordingly. However, these algorithms can often struggle to take into account all the factors that can influence the performance of an advertising campaign. This can lead to disappointing results and make it difficult to achieve the desired return on investment.</p><p>Most software claim to be using “AI” or machine learning.<br>Once again let’s be clear: <strong>Your business is NOT adapted to those technologies.<br></strong>In order to be efficient, machine learning should be run on account with huge amount of data and sales. If you make less that tens of thousand orders a month, forget “AI” solutions.</p><p>Here is an example of what an AI solution (m19) has done in 4 days on an account generating more than 50k revenue a month.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DeVpSX9SI4b43uk5B8rPIw.png" /><figcaption>380€ out of 542 spent on non converting keywords by the “AI” in less than a week</figcaption></figure><p>As the tool does not have enough data, it tries a lot of keywords, mostly irrelevant.</p><h4>How should PPC automation work to be effective for most sellers on Amazon?</h4><p>The key is to strike a balance between automation and human decision-making. Rule based automation can be a very powerful for optimizing campaigns and saving time, but it is important to ensure that there is also a human element involved in the process. This can involve combining automation with rule based advices and explanations. This way you, as seller and best person knowing your business and market would get the best insight to make the final decision.</p><p>Well good news for you. That’s exactly what we are building !<br>A tool that will give you just the right insights and advices to make the right decision.<br>That will let you know each week how your business is evolving and that will let <strong>you </strong>enable human intelligence based automation !</p><p>You can join the discord server to claim your beta invitation code.<br><a href="https://discord.gg/5HRR7R87Yy">https://discord.gg/5HRR7R87Yy</a><br>We are opening a new 30 free beta user batch !</p><p>🎁 Merry Christmas</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gwJ-m-XTG_dGrrX88_Z3Fg.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a8d2e2b2d9c3" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why you will not make money with ChatGPT (written by ChatGPT)]]></title>
            <link>https://medium.com/@nassuf/why-you-will-not-make-money-with-chatgpt-written-by-chatgpt-81acea6eed2c?source=rss-309baa13bdc3------2</link>
            <guid isPermaLink="false">https://medium.com/p/81acea6eed2c</guid>
            <category><![CDATA[make-money-online]]></category>
            <category><![CDATA[saas]]></category>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[chatgpt]]></category>
            <dc:creator><![CDATA[Nassuf Mmadi]]></dc:creator>
            <pubDate>Thu, 22 Dec 2022 13:47:04 GMT</pubDate>
            <atom:updated>2023-01-20T23:35:21.049Z</atom:updated>
            <content:encoded><![CDATA[<iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FyIxNOXEMpqkqA%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2FyIxNOXEMpqkqA%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2FyIxNOXEMpqkqA%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="322" frameborder="0" scrolling="no"><a href="https://medium.com/media/f15b33ef57874022bb3d880c7d25d4e0/href">https://medium.com/media/f15b33ef57874022bb3d880c7d25d4e0/href</a></iframe><h3>So you are considering making money with chatGPT</h3><p>While chatGPT can be a powerful tool for generating text, it is important to understand that it is not a magic solution for making money. In this article, we will explore what chatGPT is, how it can be used to create content, and why most people will not succeed in making money with it.</p><h3>What is chatGPT?</h3><p>For those living on the moon who may not have heard of it, chatGPT is a software tool that uses advanced machine learning algorithms to generate text based on specific prompts. It works by analyzing large amounts of data and using that information to generate text that is similar to the inputted prompt. ChatGPT can be used to generate a wide range of content, including articles, social media posts, and even complete websites.</p><h3>How chatGPT can be used to create content</h3><p>One of the main ways chatGPT can be used is to generate content. Whether you are a blogger, content writer, or business owner, chatGPT can help you create high-quality content in a fraction of the time it would take to write it manually. For example, you could use chatGPT to generate articles for your own website or blog, generate social media posts for your business, or create content for your own products, such as ebooks or online courses.</p><h3>How to make money with chatGPT</h3><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2F3EAKxfLyMrjSBtRn4A%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2F3EAKxfLyMrjSBtRn4A%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2F3EAKxfLyMrjSBtRn4A%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="292" frameborder="0" scrolling="no"><a href="https://medium.com/media/ad608245439fe9c1784b800591f621b7/href">https://medium.com/media/ad608245439fe9c1784b800591f621b7/href</a></iframe><p>While chatGPT can be a useful tool for generating content, it is important to understand that it is not a guarantee of success. In order to make money with chatGPT, you will still need to put in the time and effort to promote your content and build an audience. Here are a few examples of how you can use chatGPT to generate income:</p><ul><li>Generate content for your own website or blog: If you have your own website or blog, you can use chatGPT to generate high-quality content that you can monetize through advertising or sponsorships.</li><li>Generate content for clients: As a freelance writer or through a content writing service, you can use chatGPT to generate content for clients in a variety of industries.</li><li>Generate content for your own products: You can use chatGPT to generate content for your own products, such as ebooks or online courses, which you can then sell through your own website or on platforms like Amazon.</li></ul><h3>Why most people will not succeed in making money with chatGPT</h3><p>While chatGPT can be a useful tool for generating content, it is important to understand that most people will not succeed in making money with it. One reason for this is the time and effort it takes to promote your content and build an audience. In a crowded online market, it can be challenging to stand out and attract readers, especially if you are competing with other content creators who have more experience or a larger following.</p><p>Another reason why most people will not succeed in making money with chatGPT is the importance of human creativity. While chatGPT can generate text based on a prompt, it is still limited by the data it has been trained on. This means that it may not be able to come up with truly original ideas or perspectives. In order to truly stand out and create content that resonates with readers, it is important to bring your own unique perspective and creativity to the table.</p><h3>Importance of human creativity and putting it into work</h3><p>As mentioned above, the importance of human creativity cannot be overstated when it comes to creating content that resonates with readers. While chatGPT can be a useful tool for generating text, it is no substitute for hard work and human creativity. In order to succeed with chatGPT, it is essential to put in the time and effort to research your topic, come up with original ideas, and craft your content in a way that engages and informs your audience.</p><h3>Conclusion</h3><p>In conclusion, while chatGPT can be a useful tool for generating content, it is not a magic solution for making money. In order to succeed with chatGPT, you will need to put in the time and effort to promote your content and build an audience, as well as bring your own unique perspective and creativity to the table. While chatGPT can be a useful tool, it is no substitute for hard work and human creativity.</p><h3>Human conclusion :)</h3><p>This article was fully written by chatGPT with this prompt.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VfMae4OFAKPRCH0iqgNHGg.png" /></figure><p>It’s the buzz of the moment and I have been discussing a lot about it with friends.<br>What I noticed is that “while chatGPT” 🙃 can be very powerful for content and even code creation, it result depend a lot of what you ask. You can be very precise on what you are asking. Most people will have fun with it. Few will dig deeply into it. Even fewer will try to put things online, build things with it and with openAI and I think that’s why you mostly will not make money with ChatGPT. Not because it is limited, but because you still have to invest your time and put into work and most people will not.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=81acea6eed2c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to filter on multiple fields with API Platform and MongoDB]]></title>
            <link>https://medium.com/@nassuf/how-to-filter-on-multiple-fields-with-api-platform-and-mongodb-4a9ac551c97?source=rss-309baa13bdc3------2</link>
            <guid isPermaLink="false">https://medium.com/p/4a9ac551c97</guid>
            <dc:creator><![CDATA[Nassuf Mmadi]]></dc:creator>
            <pubDate>Wed, 03 Aug 2022 16:42:00 GMT</pubDate>
            <atom:updated>2022-08-03T16:42:00.291Z</atom:updated>
            <content:encoded><![CDATA[<p>I was working on a pretty simple use case on a SaaS for a client, and what should have take 20 minutes took me hours !</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FOrY1HwBBZv6DU0p3Xh%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2FOrY1HwBBZv6DU0p3Xh%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2FOrY1HwBBZv6DU0p3Xh%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="435" frameborder="0" scrolling="no"><a href="https://medium.com/media/2bd0e6889c1ad5724d708f6af6db0705/href">https://medium.com/media/2bd0e6889c1ad5724d708f6af6db0705/href</a></iframe><h3>Filter on multiple fields using API Platform and MongoDB</h3><p>We have a Product entity, actually a document as we are using MongoDB. We want to be able to perform a simple search on 3 fields of this document, using one parameter.</p><p>So:</p><ol><li>A user calls an endpoint like <strong>/api/products?search=my stuff</strong></li><li>The API searches on 3 text fields (title, brand and description) performing an equivalent of a SQL <em>“LIKE %my stuff%”</em></li></ol><p>In this case, we are not using a search engine like Elastic Search as we just want to do a first POC.</p><h3>Create a custom filter</h3><p>Let’s get straight into it: there is no native API Platform filter that can perform a search across multiples fields. We have to create our own one.</p><p>First, create a custom filter, let say “MultipleFieldsSearchFilter” in App\Filter.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*D6Zd1RhoAAu9DnjcvLBOaQ.png" /></figure><p>This filter must extend <em>ApiPlatform\Core\Bridge\Doctrine\MongoDbOdm\Filter\AbstractFilter.</em></p><p>The filterProperty() method will apply the right filters on the $aggregationBuilder.<br>getDescription() method will be used to describe the filter so it will be well documented on swagger.</p><h4>Filter</h4><p>First of all, we want to apply the filter only if it concern the right parameter and if there is a value in it.</p><pre>if (null === $value || $property !== $this-&gt;searchParameterName) {<br>    return;<br>}</pre><p>Now let’s add the right filters on the aggregation builder:</p><pre><strong><em>// Start a group match</em></strong><br>$aggregationBuilder = $aggregationBuilder-&gt;match(); </pre><pre><strong><em><br>// For each defined properties, we add a or clause <br>// equivalent to SQL &#39;LIKE &quot;%value%&quot;&#39;</em></strong></pre><pre>foreach ($this-&gt;getProperties() as $prop =&gt; $v) {<br><strong>    // &#39;i&#39; option to make it insensitive</strong><br>    $aggregationBuilder-&gt;addOr(<br>        [$prop =&gt; new Regex($value, &#39;i&#39;)] <br>    );<br>}</pre><p>That’s it ! From know you can apply your filter to any document you want with the ApiFilter annotation.</p><pre><strong><em>* @ApiFilter</em></strong><em>(<br>*      MultipleFieldsSearchFilter::class,<br>*      properties={&quot;name&quot;, &quot;brand&quot;, &quot;description&quot;} <br>* )</em></pre><h4>Description</h4><p>It’s now time to define the description so it will be well documented on swagger.</p><pre><em>/** {</em><strong><em>@inheritdoc</em></strong><em>} */<br></em>public function getDescription(string $resourceClass): array<br>{<br>    $props = $this-&gt;getProperties();<br>    if (null===$props) {<br>        throw new InvalidArgumentException(&#39;Properties must be specified&#39;);<br>    }<br>    return [<br>        $this-&gt;searchParameterName =&gt; [<br>            &#39;property&#39; =&gt; implode(&#39;, &#39;, array_keys($props)),<br>            &#39;type&#39; =&gt; &#39;string&#39;,<br>            &#39;required&#39; =&gt; false,<br>            &#39;swagger&#39; =&gt; [<br>                &#39;description&#39; =&gt; &#39;Search on the specified properties. Results must be match on at least one of them.&#39;,<br>            ]<br>        ]<br>    ];<br>}</pre><h3>The end.</h3><p>It took me a while to figure out how to search on multiple fields with mongo DB. I unfortunately did not find a well explained enough documentation for grouping “or” clauses.</p><p>I hope those simple 3 lines will help you guys 👋</p><pre>$aggregationBuilder-&gt;match()<br>                   -&gt;addOr([&#39;title&#39; =&gt; new Regex($value, &#39;i&#39;)])<br>                   -&gt;addOr([&#39;name&#39; =&gt; new Regex($value, &#39;i&#39;)]);</pre><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4a9ac551c97" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>