@josip_lebesgue Congrats on releasing a great app! Your tool will allow you to invest less but earn more Lebesgue: AI CMO🔥

@josip_lebesgue Sounds like you've harnessed a wealth of experience to create a tool that fills a real need in the marketing space. Understanding competitor moves is vital, and it's interesting to see how your platform translates that into actionable insights. How granular can we get with the ad insights? Can we see the engagement metrics on competitor ads as well?

@mocarg Hey Momcilo, at the core of our tool, we've advanced our LLMs specifically for estimating competitor ad performance. The system provides a comprehensive breakdown with various metrics, such as content type, textual elements, sentiment analysis, and the duration of ad campaigns. A handy tip I rely on: if an ad has been active for over 20 days, it’s usually a solid performer. Regarding my experience, I've managed over 50 projects and have directly invested more than $20M in Facebook ads. The insights and solutions I needed during these campaigns have been built into our software. This experience-driven approach is paying off for our users, with many witnessing a significant 40% increase in ROI. Hope that sheds some light on how we operate and the tangible benefits our tool offers! Best, J

@vitali_krupenka thanks a lot for your feedback. let us know if you have any thoughts or questions on how to use our product.

@mogabr thanks a lot - let us know your thoughts on our AI CMO

@dragocrnjac Great question! I understand the skepticism, as 'AI' can often be used more as a marketing term than a reflection of actual technology at work. In our case, AI is a core component of our product and not just a buzzword. Our AI implementation includes: 1. Forecasting: Lebesgue utilizes Prophet and ARIMA models to forecast ad performance up to 60 days in the future with high confidence 2. GAM Models to content performance and predict ROAS on different spending levels 3. Our Anomaly detection uses exponential smoothing 4. Budget split generators use multivariant linear regression 5. We've developed our own LLMs on BERT architecture to estimate ad performance 6. We have GPT-4 integration on the top of our unique marketing dataset 7. We use bots to scrape publicly available ads of over 15,000 businesses Our flagship technology, 'Next Steps', leverages all these insights. Our proprietary algorithm determines 10 to 15 actionable strategies each week to maximize your ROI efficiently. For more technical details, @tatjana_ramljak1 or @andrijana_brkic are more than happy to share further insights! What do you think, @dragocrnjac - does this sound like AI, or not?

@tatjana_ramljak1 @andrijana_brkic @josip_lebesgue damn, that definitely sounds like AI 😄 can you tell me more about why you had to build your own LLM? Was gpt-4 not enough?

@dragocrnjac Hello, Drago! We've developed our own Language Model (LLM) designed to forecast ad Click-Through Rate (CTR) by leveraging various factors. Our model analyzes ad copy alongside additional features like target audience demographics, the average order value of the shop, and specific product attributes. This comprehensive approach enables us to make more accurate predictions and optimize ad performance effectively. 😄

@kate_ramakaieva Thanks Kate, appreciate the support! Lebesgue works great for all industries! eCommerce is one of our stronger sides, but we've seen our users get great results in multiple different industries and I can confidently say that we're a good fit for everyone. :)

Hey @kate_ramakaieva, to build on Toni's insights, our approach – which involves tracking competitors' strategies across Facebook, Google, and email, as well as monitoring global trends – is versatile enough to benefit any industry. Currently, it's particularly effective in the fashion, jewelry, food, and beverage sectors, mainly because we've extensively scraped data from numerous businesses in these areas. However, the methodology is designed to be universally applicable, offering valuable insights regardless of your specific niche. Thank you for your feedback and kind words - definitely try our tool, it offers a 14-days free tria, and let us know your thoughts :)

@nicolaas_spijker thank you for your kind words on our launch! It's great to hear from someone who values what we're building. Here's a quick look at what we're rolling out soon: 1. Competitor Email Strategy Tracking: Get insights into how your competitors approach their email marketing (campaigns, flows, and frequency of sending emails). 2. Enhanced LLMs for Ad Performance: Improved architecture for higher precision of our ad performance estimators. 3. Promotional Ad Filtering: Easily distinguish between promotional and non-promotional ads of competitors weekly. 4. Creative Strategist Tool: We're particularly excited about this one! It's designed to offer comprehensive feedback on your creatives and targeting. Using a wealth of parameters and industry insights, this tool will guide you in crafting more effective and impactful ad campaigns. Feel free to share any thoughts or needs you have! Best, J

@josip_lebesgue Wow these are all super exciting! As a marketer who's been dabbling into email lately I think the competitor email strategy tracking will be super valuable and a great add-on to the already offered functionality! #4 sounds super interesting as well, looking forward to seeing all of these live. 😇 Enjoy the remainder of your launch! :)

@nicolaas_spijker thanks a lot, I really appreciate your thoughts and feedback. Definitely continue following us, our ideas of how marketing tool should look like align nicely.

Hey @alessandra_piccioni, thanks for a great question. We're launching platform platform-agnostic version by the end of the year (@rbenic_lebesgue make sure to deliver it on time for Alessandra :) ). In this blog post you can see how you can test our product right now, even if you do not have a Shopify or woocommerce store: https://lebesgue.io/growth/how-t... (it would be good to add competitors you want to track right now and start using our software once we're not e-comm dependent)

@alessandra_piccioni @josip_lebesgue I would add that we are also planning on integrating with more platforms in the future including Adobe Commerce/Magento, Squarespace and Wix.

Hey @darya_skd thanks for a very good question! Our app collects data from all the industries our customers are part of and computes market benchmarks for every industry separately so that we would avoid the trap of comparing apples to oranges. This approach recognizes the diversity of industries, acknowledging that what may be considered good performance in one industry could be different in another. The type of our users varies a lot, they span over 11 very different industries and our algorithms are very robust and work well regardless of industry. You can check out success story of one of our customers on the following link: https://lebesgue.io/growth/ai-dr...

@darya_skd, thanks for your excitement about our launch! We've seen significant impacts in industries like fashion, jewelry, food, and beverage, where our tool's AI-powered strategies have helped businesses optimize their marketing campaigns. By leveraging competitor analysis and trend insights, these industries have particularly benefited in terms of increased ROI and more targeted ad campaigns. We're thrilled for you to try it out and see the difference it can make in your marketing efforts! 🚀📊

Hi, @martin_sosic, we track many default competitors for every industry. However, you can pick your own competitors since this is 100% customizable in the app and we’ll start tracking these straight away and scrape ad libraries and their social networks for you.

Hey @martin_sosic, thanks for reaching out with your question. Absolutely, you can add any competitor you wish to track in our system. The main thing to note is that for new additions, we start collecting data from the moment they are entered into our system. This means while you won't have immediate access to historical data spanning the last two years, the value of the insights will grow over time as more data is accumulated - a neat feature we believe helps with user retention. 😊 And for those in the e-commerce sector, we've already amassed a significant amount of data on industry leaders, providing rich insights from the get-go. Let us know your thoughts!

@josip_begic @toni_perko @dzaitzow Hey there! First off, thank you so much for your kind words and enthusiasm about Lebesgue! We're thrilled to hear that you find the Competitors Intelligence feature and our AI-driven approach as impressive as we do. Regarding our tech stack, while I can't dive into tech specifics (gotta maintain a bit of mystery, right?), I can definitely share a high-level view. We've built Lebesgue on a foundation that combines robust data science algorithms with advanced AI components. Our stack is a carefully curated mix of machine learning models (BERT Architecture for LLMs, ARIMA and Prophet forecasting) for analytics and predictive insights (exponential smoothing for outlier detection), alongside cutting-edge tools for data processing and real-time benchmarking (classic scraping tools with some added complexities :) ). This blend allows us to dynamically process and interpret large volumes of data, ensuring that our users receive timely, accurate, and actionable insights. The goal is always to make data-driven marketing not just a concept but a tangible, accessible reality for our users. Hope this gives you a little peek into our world without giving away the secret sauce! Thanks again for your interest and support. Best, Josip

@josip_begic @toni_perko @josip_lebesgue Wonderful - and I totally agree about the mystery! Curious from a costing perspective - how complicated has that been for the team to forecast spend given all the LLM architecture?

@josip_begic @toni_perko @dzaitzow Very complicated :) We're using different forecast methods and estimating what performs the best in different cases (some models are better for longer forecasts, some for sudden spikes in data). LLMs are used to estimate the performance of your competitor ads (CTR and CR) - that is a really complex problem and we can definitely improve it further :)

@josip_begic @toni_perko @josip_lebesgue Very neat - We're pre launch and don't really have a use case for this at the moment but in the future it is top of my list if a tool like this is in need!

@rishi_raj_jain_ Thank you for support. You can learn more about next steps insights feature here: . This feature is designed to prioritize your business's next moves by leveraging Competitor Intelligence and analyzing vast global marketing data.

hey @rishi_raj_jain_ , yes, insigts take into the account Facebook, Google, Email strategies and figure out next steps that you should do as a business owner to improve your results.

@gabriele_mazzola I wasn't sure should I reward this comment with LOL or the bright idea for the feature suggestion. @josip_lebesgue Car key tracking a new feature on the roadmap?

@josip_lebesgue @leon_novacki hahaha you're welcome guys! If you need any incredibly useful ideas such as this one, just reach out! Happy to contribute LOL

@scythe1 Thanks for your support Neha!

@scythe1 Thanks a lot, let us know your thoughts once you try the tool.

Thanks a lot for the awesome question and kind words @anthony_latona. Behind our AI, there's a robust combination of technologies. We utilize advanced machine learning models, including natural language processing (NLP) and image recognition algorithms. These are integral for analyzing ad content and extracting valuable insights. For the NLP components, we rely on state-of-the-art techniques to understand and categorize text elements in ads. This helps in sentiment analysis, keyword extraction, and trend identification. On the image recognition front, we use deep learning models to analyze visual content, identify themes, and assess the effectiveness of different visual strategies. Additionally, our database management and analysis are powered by sophisticated algorithms that handle large volumes of data efficiently. This includes scraping tools that collect ad data from platforms like Facebook and Google, and analytic tools that process and make sense of this data. Our system also incorporates insights and data from our extensive experience in digital marketing, especially from our $30M ad spend on Facebook and Google. This human experience, combined with AI, creates a powerful tool for understanding and predicting effective marketing strategies. We also use Generalized Additive Models, ARIMA and Prophet forecasting techniques, exponential smoothing and multivariant linear regression for various parts of our software (outlier detection, budget allocation, revenue forecasting, seasonality and trend detection, etc.) Again, thanks for the encouragement, and I'm glad you see the potential of our tool in saving money for marketers and boosting company revenues!

@sviridov Thank you for recognizing the effort and dedication that went into creating Lebesgue! It's true, developing a comprehensive tool that provides detailed insights into competitor strategies and helps businesses enhance their own marketing efforts required a lot of time, knowledge, and passion 💪. Our team is thrilled to see that our hard work is being appreciated.

@goran_golub Hey Goran, appreciate your support! 😄 Regarding the optimal distribution of the advertising budget, we've got you covered. Our app features a budget split insight generator that utilizes your data and past performance to calculate the optimal allocation between Facebook and Google. Currently in development, we're working on expanding this feature to include additional advertising channels. 🚀

@chtrbx_ai All true, Jake! Happy you like it and I hope you'll give Lebesgue a try for yourself and see how we can help you improve your marketing results!

Hey @chtrbx_ai, absolutely, that's exactly what Lebesgue offers! It might sound a bit like magic, but it's all about smart technology and comprehensive data analysis. Our tool provides detailed insights into your competitors' ads across various platforms. You get to see not just the ads themselves, but also valuable data on their performance and strategies. By analyzing this information, you can understand what's working for them (and what's not). This insight allows you to either replicate successful strategies or go one step further by enhancing them for your own campaigns. It's like having a window into your competitors' marketing departments, giving you the opportunity to learn from their successes and mistakes. Think of it as a significant leap in competitive intelligence, all designed to give you an edge in your marketing efforts. It's a game-changer for any business looking to stay ahead in the fast-paced world of digital advertising. 🌟🔍🚀

@zvonimir_sabljic1 Hey Zvonimir, no catch here 😎 Just check out our reviews on the Shopify App Store and you'll see that we've been delivering great results for our customers for a while now and you're right, it really is a no-brainer for anyone running their business on Shopify!

@toni_perko Congrats, yea, the reviews look amazing - I would definitely link that somewhere here so people can see.

@tomaslau Hey Tomas, Thanks for the support here! We've got a lot of users from other industries as well and we're not solely limited to eCommerce businesses. With the AI Copy Creator and the competitor tracking features, I'm sure you can get an edge on your competitors! Give it a try, I'd be happy to connect and help on a call too!

Hey @dragocrnjac that's a great question and not noobish at all! Understanding the dynamics of ad activity is crucial in digital marketing. Here's how it works with our system: 1. Scraping and Monitoring: Yes, we do scrape platforms like Facebook, Instagram and Google Search (Email coming soon). This process involves systematically collecting data from these platforms to keep track of various ads. Every time we detect an ad, we screenshot it and store in our system 2. Identifying New Ads: When a brand launches a new ad, our system detects it as part of our regular scraping process. We catalog these new ads and add them to our database (by comparing them with previous ads already stored). 3. Tracking Active Ads: For ads that are currently running, our system regularly checks their status. This allows us to know which ads are active at any given moment. 4. Determining Inactive Ads: When ads stop appearing in our subsequent scrapes or show signs of being paused (like no longer receiving engagement, views or just disappearing), we categorize them as inactive. This process helps us provide our users with up-to-date information on the ad strategies of various brands, giving insights into what's new, what's working, and what's been retired. Hope that clarifies things! Best, J

@lubapav you can always check our blog posts to learn about the latest strategies and best practices: https://lebesgue.io/digital-mark...