Cansu Kaya

Are you an aspiring data engineer or data scientist looking to hone your skills with real-world data? Look no further! I've just published an article where I share some fantastic open datasets that you can use to practice and refine your data engineering and data science skills. Link : https://lnkd.in/g7tZtE8a In this article, you'll find: - A curated list of open datasets across various domains. - Practical examples of questions you can try answering with the data. - Tips on how to approach data analysis and engineering projects. These datasets offer a great opportunity to apply your skills in data cleaning, transformation, visualization, and machine learning. Whether you're preparing for a job interview or just looking to build your portfolio, these datasets will provide you with the hands-on experience you need. Happy coding and analyzing! 💻📊 #DataScience #DataEngineering #OpenData #MachineLearning #AI #BigData #DataAnalysis #CareerGrowth #TechCommunity

30

7 Comments

Create great data graphics for social in 5 minutes or less - here's how: 1. Build chart in data viz program (I prefer Tableau, but there are lots of options).     2. Export chart into Figma or other design tool, insert into pre-built template.     3. Take final graphic from Figma into LinkedIn / Twitter / wherever, write something useful. Just that easy! Obviously there are many steps before you get to the data visuals themselves (data modeling, cleaning, etc) but if you're working with a reliable dataset it really is this simple. Been getting a lot of questions about the system for Insights here at Carta - happy to walk through our tooling and thought process for any weekend tinkerers. Happy vizzing! #insights

149

42 Comments

If you use the right predictive model, you can have an outsized influence on the "movable middle." Here's how our approach is moving the needle for our clients: 1. Model Selection: We're flexible. Whether it's a regression model or a sophisticated machine learning algorithm, we choose what works best for the specific challenge at hand. 2.Customer Ranking: Our models rank customers from most to least likely to exhibit a desired behavior. 3. Decile Division: We split these rankings into 10 equal groups, or deciles. From there, we can lump customers into one of three categories: -Top Deciles: They'll likely engage without extra incentives. -The Middle: This is often where the greatest opportunity lies. -Bottom Deciles: Unlikely to engage regardless of efforts. We focus on that potentially movable middle group. By testing different offers, messages, and channel mixes, we can influence behavior where it matters most. By understanding where customers fall in these deciles, we can optimize marketing efforts and allocate resources for maximum impact. #DataDrivenMarketing #CustomerInsights

9

1 Comment

Want to ditch data analysis confusion & become a rockstar? Our latest post defines reproducibility & shows you how to document your work for clear, repeatable results. https://wix.to/kEpcU5r #newblogpost #DataScience #Reproducibility #dataanalystskills #analytics

2

1 Comment

No one likes sharing the boring stuff, but this 'boring' stuff is what gives us data scientists our foundation. No matter what project I venture into, statistical distributions are ALWAYS there somewhere, and, other than the normal distribution, I often forget there are other practical, more nuanced members of this statistical family. This article is a simple (non-mathematical) refresher on some of the more popular distributions I regularly come across. ** If you don't find the descriptions useful, at least you can enjoy my simple Seaborn visualisations! #python #seaborn #matplotlib

7

With self-serve analytics, you can empower everyone in your company to go from data to decisions. Our VP of Analytics Adam Kinney explains how we do it at Mixpanel 👇 https://lnkd.in/dUeb4mDe

6

I believe in the power of data to solve big problems. I also believe that we suck at talking about it. But I also ALSO believe if we can get better at talking about it, we can have a huge impact on issues from business to politics to society and to the environment. When it comes to the problems, we’re all a part of it. Data doesn't speak for itself, and the goals of good communication don't align with good analysis. -------------------------------------------------------- Read more in depth about the challenges of communicating with data, how we can overcome them, and some exercises to help you improve your own communication in the Penguin Analytics newsletter. #analytics #datascience #communication #newsletter

3

🛑 ❗ Leaving technical names in your dashboard smacks of inconsideration for your audience. 👀 I’ve seen both inexperienced and more seasoned dashboard developers guilty of this underscore-toting offence and it makes me cringe every time! 😬 You can’t expect your business users to correctly guess what each graph is showing from alphanumeric-and-underscore soup. 🍜 Even for fairly straightforward terms like sales_num - the mature dashboard designer will ensure it is changed it to a properly capitalized and spaced business name label ---- before it ever rests on business users' eyes! 👩‍💼 👨‍💼 Check out my article for more info 👇

11

1 Comment

Looker Studio - #Funnel blog posts   Discover or rediscover 3 tutorials about different solutions to create funnel visualizations inside #LookerStudio.   1️⃣ The Native - A recent post helping to adapt your data set to fit with the freshly released funnel chart type. 2️⃣ The Scalable - A tabular viz to build a tunnel with a single table chart & all the data available in one view. An adapted solution when there are a lot of funnel stages. 3️⃣ The Interactive - To give the control to the report readers to display what they need as unique metric for a simple viz.   3 years, 3 articles, so I would not be surprised to write an new article in 2025 about designing funnel...again...😊    The link to these articles is available in the comments section.

37

2 Comments

For those who want a 5 min crash course on MMM and potential biases that can impact modeling, this is an incredibly useful read.

2

If you're struggle to quick measure your Tableau workbook speed you should definitely check out tableau-speedtest.site. It's free and open-source. Also I wrote an article on why do you might need to use it: https://lnkd.in/dVYuRgjk To contribute go to GitHub and DM me for further information! Have a great Tuesday, #DataFam! #tableau #performance

47

3 Comments

🧩 Reading [Google Research's latest article](https://lnkd.in/ejk3_wMH) on understanding relationships between datasets got me thinking about the bigger picture in data management. This research brings out the importance of uncovering and automating complex relationships within raw data—a task that’s essential but often challenging. ✨ Palantir’s Ontology approach came to mind here as a powerful example of putting these ideas into action. By creating a digital twin of business entities and their relationships, Palantir enables companies to interact with data at a higher, business-relevant level. This abstraction layer transforms raw data into meaningful insights. ⚙️ To bring further structure, tools like Featuretools, Woodwork, Compose, and EvalML offer an open-source framework set that standardizes features and prepares data for machine learning. With **Featuretools** for automated feature engineering and **Woodwork** for consistent data typing, this toolkit enables the creation of reusable data products right from raw schemas and tables. 🌐 Imagine combining Google’s research-driven insights, Palantir’s digital twin framework, and the Featuretools ecosystem. Together, they could enable a truly dynamic data ecosystem—one where relationships are not just mapped but also standardized, enriched, and primed for decision-making. #DataManagement #Ontology #FeatureEngineering #GoogleResearch #Palantir #MachineLearning #OpenSource

1

If you want to design better dashboards and more impactful data products, the biggest upgrade won't come from learning more about a tool. Learn more about the business and learn more about your stakeholders. Here are some ways you can do that: - Spend some time working alongside them, or spend a day shadowing them to understand what they do and how they operate. - Sit in on critical meetings where your data is used. Listen and learn how they talk about the business. - Have coffee with your stakeholder's stakeholders. What are their pain points? What are your stakeholders sharing and not sharing? - Spend some time with your finance peers learning how money flows through the business. Where does it touch your role? A deeper understanding of your stakeholders and business will focus your dashboard on the things that matter. --------------------------------------------------- Want to be a rockstar analyst? Learn to talk about data better with free communication tips every Sunday:  https://lnkd.in/eJYBuVPQ

2

Today I want to ask Tableau about divergence between web edit and desktop. There are so many nice features in web edit I wish were in desktop as well. E.g. - Shape size percentage guage - slide shape size with precision. - Hyper links and spell check in text boxes. - Layout hierarchy for dashboards - move sheets up or down into the desired container with ease. Yet web edit doesn’t have the full formatting functionality that desktop does, to make the full switch. Switching between desktop and web is frustrating to get the full range of functionality. Tableau driven by agent force is a brand new UI coming next year. This is an opportunity to converge the product interfaces. For a better seamless workflow between desktop and Cloud. My question for Ryan Aytay would be is there plans to converge or continue to diverge web edit and desktop? #tableau #webedit #desktop

8

2 Comments

Your Data Infrastructure (or lack thereof) Is Costing You BIG TIME This simple adjustment led to 110% more purchases and a 33% reduction in Cost Per Purchase. What was the change? Was it new creatives? Nope. Was it new copy? Nope. Was it simplifying their offer? Nope. The real game changer? Hiring our team of specialists to optimize the D A T A infrastructure & strategy Your Data Infrastructure is the foundation of ALL your marketing and growth activities & a robust first-party data strategy is crucial for your growth efforts. Here's the issue. . . Marketing and growth teams are stretched thin— managing clients, creating content, testing creatives, developing strategies—the list goes on. And what inevitably happens is that the very foundation of all of their efforts is left unmanaged. The consequences of neglecting your data infrastructure and data strategy include: > Erratic ad performance > Rising acquisition costs > Missed growth opportunities All of which directly impact your bottom line. After the release of iOS 14.5, brands have adopted a variety of strategies to overcome tracking and data challenges, typically falling into three main categories: Bucket 1: Ignore the need for effective first-party data systems & endure rollercoaster ad performance. Bucket 2: Use standard CAPI through platforms like Shopify. Better than nothing, but still limited and will never deliver optimal performance. Bucket 3: Rely on tracking, CAPI or attribution tools just because a well known podcast or thought leader recommends it. Spoiler alert. . .most of these tools are just glorified tag managers. Most of our clients come to us from Bucket #3—using tracking, CAPI, or attribution tools recommended by influencers or podcasts. When they put these popular solutions to the test against us, we've never been outperformed. Upgrade your data strategy today with audienceOS & join our clients who experience: ✅ More consistent ad performance ✅ Lower costs per acquisition ✅ Enhanced ability to scale Bottom line, an optimized data strategy & infrastructure are no longer 'nice-to-haves'—they’re essential for consistent, predictable growth. Don’t let your data infrastructure be your Achilles’ heel. Invest in your data infrastructure today & unlock massive growth and stability for your brand or clients. Comment below or send me a DM & let's see how much more profit we can help you squeeze out of your paid acquisition efforts.

3

Google’s decision to postpone the deprecation of 3rd-party cookies is unsurprising... And I hate it. 3rd-party cookies are the boiling frog of our industry. There are few worse ways of measuring performance than relying on the shaky data foundation of 3PD. But it’s easy data to access, so companies keep defaulting to it, despite its value constantly eroding. Last week, Dr Grace Kite shared that Last-click attribution costs advertisers £74bn a year in lost sales in the UK & US. That’s equivalent to ~24% of all digital advertising spend. Yet those marketers will now go back to twiddling their thumbs. They will talk about “the AI revolution” and ignore that good data is the bedrock of all they hope to achieve. Or in other words, being boiling frogs. But oh well... There is always next year.

47

6 Comments

Malcolm M. Bonner Right on! I think it's misguided to believe that everything worth knowing and expressing can be captured solely from the data. The root of the issue is the purpose of the analysis. If an analyst is using the model to simply summarize the relationships expressed in the data, then they're best to only use the data. But who want's that from a marketing mix model? If you think so, then just question why it's important to iterate on the specification of the model so much. The unspoken truth about marketing mix is that typical analysts iterate to no end to ensure they get "the right" answer. The problems with marketing mix models come from the data, not the methods. The problems are collinearity, low variance and sparsity. They are the direct result of management actions to drive the business. The data cannot be relied as the exclusive source of information to inform such important matters as marketing effectiveness. In my view, marketing mix models are a decision-making solution and so they need to be working with the full knowledge estate. The knowledge estate of the marketers as decision makers. This includes but is not limited to all the sources of information listed by Malcolm M. Bonner. For a more in depth discussion on this topic, please see my post on https://lnkd.in/eMvrTFkU

4

1 Comment

We have been BUSY 🏃‍♀️ Check out our latest Product Hunt launch of our new RAG magic 🧙‍♀️ 🤖 Go from dataset to custom RAG in 5 minutes, no code needed 💁‍♀️ https://lnkd.in/gPXdF-ZQ #AI #Technology #GenAI

10