GrowthBook’s Post

Here's 101 on "Software Industry vs. Analytics Industry." We have been using softwares for a very long time. The software industry uses nearly all of the best practices associated with each & every stage of the software development lifecycle. You name it, team collaboration? version control? integration with third-party tools? documentation? testing? CI/CD? It's already there. But let's consider the analytics domain for a while. ❌ Have you ever been able to collaborate on the dashboarding process? Like, can two people work on the same dashboard at the same time? NO ❌ Can you get automated documentation of your pipelines/dashboards?NO ❌ Can you version control your SQL queries/dashboards?NO ❌ Can you automate the testing of your pipelines/dashboards?NO The "Analytics industry" has been craving the best practices of the "Software Industry" for a long time. Analytics workflows don't have end-to-end version control, documentation, testing, CI/CD, and lineage all on one single platform. Analytics folks have to use different tools for different features. Sure, there are ways to use stored procedures, save dashboard's xml/json files in git, and other hacky ways to save your work and collaborate (to some extent), but these aren't production-ready solutions. These are the pain points that new-gen platforms like Microsoft Fabric, DBT, etc. are trying to solve. ✅ In one line,they are trying to bring structure to the analytics industry. Want to learn how to create a data pipeline using dbt? I am hosting a free workshop on Sunday, 28th April, at 12PM IST Here's the link to workshop: https://lnkd.in/djwD4Q8b Shashank Mishra 🇮🇳 Rahul Shukla

I was bored building APIs until this application hit my plate. 5 lessons I learned from building a Data-Driven application. The Request: Build an application to help customers visualize their data. It has to be fast and follow our design guidelines. 1. Enterprise Data platforms are expensive; look around. The obvious solution was to look for a platform to connect to the data and build some visualizations. I looked at some enterprise solutions, and GOD, they cost an arm and a leg. I needed a plan B, so I took a deep dive into the open-source pool. (This pays off 90% of the time) I found two main options: Taipy and Streamlit.  2. Data Integration doesn’t need to be painful. One of the first real headaches was trying to get all sorts of data to play nice together. There was no unified view. The data live across different formats and Systems. That was when I discovered Taipy’s data dashboards. Now, I can bring together data from different sources: - Databases - APIs - Real-time feeds Also, Data Dashboards serve to simplify and make complex data visually attractive. 3. Staring at numbers is not enough. Users want to ask "what if" and watch the data come alive with answers. I need to develop something interactive that can provide quick responses. Again, Taipy provided more out-of-the-box features for creating interactive GUIs without sacrificing simplicity. 4. Your users deserve nothing less than the best UI. Taipy's capability to build responsive and interactive graphical GUI was a game-changer. Streamlit renders all the graphical components again when a user interacts. But, when the user interacts with a visual element, Taipy triggers callbacks based on the action or change in the GUI. This was a decision maker; check how smooth the transition is on the image.  5. Users want data, but only if it is fast. No matter how fun your spinner animation is, nobody wants to spend time looking at it. As the platform grew, so did the concern about its scalability. Taipy's scalable architecture allowed me to manage the growing demands of the platform. The open-source community came through, and I can’t sing Taipy’s praises enough for making my job much easier. Give them a star here: https://lnkd.in/eYsaZGEr And start today by typing: $ pip install taipy Big thanks to Taipy for supporting this post.

📈 𝐅𝐫𝐨𝐦 𝐒𝐭𝐚𝐭𝐢𝐜 𝐑𝐞𝐩𝐨𝐫𝐭𝐬 𝐭𝐨 𝐃𝐲𝐧𝐚𝐦𝐢𝐜 𝐈𝐧𝐬𝐢𝐠𝐡𝐭𝐬: 𝐀 𝐃𝐚𝐭𝐚-𝐃𝐫𝐢𝐯𝐞𝐧 𝐒𝐮𝐜𝐜𝐞𝐬𝐬 𝐒𝐭𝐨𝐫𝐲 We're excited to share our latest customer success story, focused on the commercial real estate world and the transformation journey required to support their business intelligence capabilities using #Microsoft technologies ✨. Find out how they overcame challenges such as limited visibility and integration by leveraging a tailored solution that enabled them to access real-time #data and insights across their organisation. Get the full low down here 👇 https://ow.ly/pRsO50SAj4V #datadriven #customersuccess #togetherwewin

𝙄𝙢𝙥𝙡𝙚𝙢𝙚𝙣𝙩 𝙮𝙤𝙪𝙧 𝙩𝙤𝙤𝙡𝙨 𝙬𝙝𝙚𝙧𝙚 𝙮𝙤𝙪𝙧 𝙪𝙨𝙚𝙧𝙨 𝙚𝙭𝙥𝙚𝙘𝙩 𝙩𝙝𝙚𝙢 𝙩𝙤 𝙗𝙚 If you want your users to use the tools you are building or onboarding, it’s important to implement them exactly where they expect them to be. You’re probably expecting a data contracts example, so for variety I’ll use a data governance one instead :) It’s often a requirement to categorise our data, so we can track what personal data we have, how long we’re keeping it for, how we’ll anonymise/delete it, and so on. Much of that data is owned by product engineering teams, who produce and manage that data through the code they write. It would be asking a lot of them to categorise the data in a data governance tool through a web UI, or in a spreadsheet, or some other interface they wouldn’t go to otherwise. You might get them to do it once or twice, but sooner or later it’s going to be forgotten about and fall out of sync. Instead, can you allow them to categorise the data in their code, as close as possible to where they are defining their data models? If you do that, you could also provide them with continuous integration checks to ensure they are completed and prevent common mistakes. If you have a requirement to have this presented in your data governance tooling, you could ingest this data into it. If you want your users to use your tooling, implement them where they expect them to be. #DataContracts #DataPlatform

🔥 Embedded analytics integration can be challenging, but it doesn’t have to be. In our latest blog, we dive into the top 5 best practices to help you overcome integration hurdles, from compatibility issues to data consolidation. 🚀 Learn how Reveal can make your integration seamless. Key takeaways: ✅ Leverage robust APIs and SDKs ✅ Focus on scalability and performance ✅ Deliver a smooth user experience #EmbeddedAnalytics #DataIntegration #Scalability #ProductDevelopment

💡 Leveraging Custom Software for Better Data Management and Analytics 📊 I’ve been reflecting on how custom software changes the game for businesses, especially when it comes to data management and analytics. Data is the new gold, but what good is it if we can’t manage or analyze it effectively? 💻💡 Custom software solutions are like a tailor-made suit for your data needs – perfectly fitted to your organization’s unique requirements. Instead of forcing a generic tool to fit, why not create one that molds to your processes? Here’s what I’ve found: ✅ Enhanced Data Integration: With custom software, you can seamlessly pull data from multiple sources. No more switching between platforms or dealing with messy integrations—everything flows into one place, making data analysis smoother and more reliable. 🔗📊 ✅ Real-Time Analytics: Custom tools are not just about data; they deliver insights exactly when you need them. Imagine receiving real-time reports without delay, empowering you to make swift, informed decisions. It’s a game-changer! ⏱️📈 ✅ Security and Compliance: In today’s world, data security is a top priority. Custom software can be designed with built-in compliance checks and security measures tailored to your industry needs. Sleep better knowing your data is secure! 🔒✅ ✅ Scalability at Its Best: As your business expands, so does your data. Custom solutions are designed to grow with you, ensuring you’re never caught off-guard by unexpected data growth or changing requirements. 📊🚀 ✅ User-Friendly Interface: One of the best things about custom software? You have a say in the user experience. No more complicated dashboards or overwhelming data sets – just a clean, intuitive interface designed for your team. 🎛️👌 From my experience, investing in custom software has been like unlocking a superpower for our data strategy. How are you managing your data? Have you tried custom solutions, or are you sticking with off-the-shelf tools? Would love to hear your thoughts! 💬👇 #CustomSoftware #DataManagement #BusinessIntelligence #RealTimeAnalytics #TechSolutions #DataIntegration #Scalability #DataSecurity #AI

In the captivating realm of #data-driven innovation, #recommendation systems promise to revolutionize businesses by supposedly understanding customers' needs like a sixth sense, leading us through a riveting journey from data collection to algorithmic enlightenment.

People overcomplicate embedded data portals. They believe users want ultimate control. I don’t see it. The last few years has shown huge growth in embedded analytics use cases as organizations look for new ways to 1/ Create visibility to external partners 2/ Build data transparency with community 3/ Monetize a new revenue stream with reporting The deployment considerations are a little more complex vs internal use cases. - Security is even more important - White-labeled branding and UX are critical - Building what customers need vs want is nuanced The last point is interesting. People assume users want the ability to create their own dashboards - supposedly power users with too much extra time on their hands. I don’t believe it. The issue with external self-service is how complicated the path of designing, saving, and revisiting ad-hoc reports works. And despite the investment to build the flexibility and functionality, rarely do I see external customers build their own dashboards. People don’t have time. People expect your reports to fit needs. Don’t overcomplicate external reporting portals - start with well-defined dashboards that solve to the 80%. Test, iterate, and partner with your heavy users. Way easier than assuming they’ll build for you.

Struggling with data import and transformation? Look no further! Data Fetcher makes it easy to import data from various applications and APIs into Airtable, no coding required. Plus, enjoy features like real-time financial data, custom API requests, and team collaboration. Check it out now! Join our newsletter for more tech tools like DataFetcher and stay ahead in the game: https://lnkd.in/gev_UBwT #DataTools #DataImport #Airtable

Unleash the power of 'Send Web Request' in AITable.ai! 🚀 Our latest blog shows you how to automate cross-table data sync effortlessly. Dive in now! 📊 https://lnkd.in/g-JbKJMA #DataSync #Automation #WebRequest