Goodbye 2025 🎄 - Issue 297

A recap of 2025, confetti, gratitude, and what’s coming next!

Welcome to the last newsletter of the year 🎊 from the Data Analysis Journal, a weekly newsletter about data science and analytics.

If you missed December’s posts, here’s the roundup:

• Refresher on Experimentation - A consolidated guide to statistics, tools, best practices, and advanced experimentation.

• Refresher on Visualizations - Charts, dashboards, and storytelling basics - what actually matters when showing data.

• Refresher on Metrics and KPIs - An evergreen guide to the metrics you’ll use for experimentation, analysis, modeling, and forecasting.

Another devastating year for Ukraine

Today marks 3 years and 10 months since the full-scale war began in my home country.

I don’t usually write about the war here. I know that for many readers it feels far away. But this war has directly impacted me, my family, and many of my friends and colleagues. My home was wiped out. My school was destroyed. The church I was married in was bombed. Watching what Russia is doing to my homeland is devastating.

It’s also painful to see that 86% of Russians support this war. So once again, I’m asking you to take a clear stance:

• Don’t buy or use Russian products or services.

• Refuse donations, funding, or sponsorship from Russian organizations or affiliates.

• Decline cooperation with Russian artists, writers, bloggers, engineers, and scientists.

• Refrain from promoting Russian art, poetry, culture, and cuisine.

Russia could stop this at any time. They choose not to.

If you know Ukrainians who have recently relocated, please support them however you can. There are about 6.9 million Ukrainian refugees globally. Starting life over in a new country is brutally hard - it’s 10 times harder to find a job, make friends, or feel a sense of belonging.

Proceeds from newsletter subscriptions are donated to the Ukrainian Children’s Aid and Relief Effort (UCARE) charity. I am on the UCARE board, and we’re raising funds for food, clothing, medicine, and basic necessities for children impacted by the war.

To the Ukrainians reading this: stay strong. As always, I offer my newsletter, mentorship, and support at no cost.

⏪ A rewind of 2025

This newsletter reflects my own journey in analytics. As I took on new projects and grew my team (we now support 24 products and apps), I also pivoted the content toward the problems I’m solving every day. This year, I stepped back from SQL/Python tutorials and one-off EDAs to double down on subscription analytics, ROI measurement, forecasting, and critical thinking.

Most analysts can write SQL and create a plot or a dashboard. The hard part is stitching the puzzle pieces together: finding the “so what,” connecting metrics, and turning analysis into a decision. That’s what I focused on in 2025. Here’s a roundup of the most popular publications from my newsletter over the last 12 months:

Set your foundation

• How To Pull Stories From Data and Make Them Actionable

• 10 Must-Know Concepts Every Analyst Should Know

• How to Create Effective Decks

• Structuring Data & Analytics Teams for Real Impact

Analytics in action

• Analysis for Optimal Cadence and Frequency

• Forecasts Part 1: Choosing the Right Approach

• Forecasts Part 2: Building Forecasts

• From Forecast to Action

• How to Analyze Onboarding Flows

• Why Your Activation Analysis Is Wrong - And How to Fix It

• Inside Product Analytics: Decoding User Behavior Part 2

• Re-Subscribers: Are Your Customers Coming Back?

A/B Testing

• Significance Level vs. Statistical Power in A/B Testing

• Rethinking A/B Testing for B2B and SaaS

• How Notion Scaled Experimentation | Ekanth Sethuramalingam

• 10 Experiments Every Data Team Should Run

• Refresher on Experimentation

Measurement and Reporting

• Why Trial Success Rate Is the Hardest KPI To Accurately Report

• Why Most Benchmarks Are Misleading - and What to Use Instead

• How to Report on DAU/MAU Ratio

• The Three Layers of Retention Analytics

• Upgrades and Downgrades: Where Most Reporting Goes Wrong

• Why Grace Periods Keep Screwing Up My Reports (and Probably Yours Too)

• When Simple Becomes Tricky: Should Discounts Be Included in MRR?

• Churn: Why Most Teams Get It Wrong

Career growth, advancement, and self-learning

• How To Differentiate Yourself in Analytics and Stand Out

• How To Develop Analytical Intuition

• When Great Work Isn’t Enough

• Which Analyst Are You?

• What Is the Best Advice You Have Ever Received?

Tooling / Tech

• Tableau vs. Power BI – Which Should You Choose?

• BI for Builders: How Superset Stacks Up in Modern Analytics

• How To Set Up Analytics for Web and Mobile Products

• Semantic Layers: The Right Idea, the Wrong House

🌟 Thank you to my guest writers or presenters for sharing their expertise this year:

• From Numbers to Narratives | Chris Dalla Riva

• Causal Inference Methods for Bridging Experiments and Strategic Impact at Roblox | Wenjing Zheng

• A New Chapter for Mixpanel: My Conversation with CEO Jen Taylor

• Reinventing How Analysts Work | Barry McCardel

🚀 What’s next in 2026

AI is accelerating team workflows. For us, that means more apps, more tools, and more projects - which increases the demand for strong analysts. Teams are realizing that scalable growth depends on our ability to connect users, product features, and revenue, and to make sense of the whole system. If you can’t unify signups, transactions, and product engagement data, you can’t use AI effectively - you can’t automate, and you can’t even trust what your analytics are telling you.

So in 2026, I’ll be focusing more on:

1. Tooling and tech for analytics - how to unify analytics workstreams, map users across systems, capture the context AI needs, and maintain these pillars as you scale.

2. The foundation of analytics - avoiding busywork by asking better questions, thinking critically, challenging assumptions, and turning scattered data points into a clear narrative.

3. Experimentation - I don’t know a company that doesn’t A/B test. Experimentation frameworks should be owned by analysts, and I’ll keep covering the fundamentals, pitfalls, and practical guidance for running reliable tests.

🏆 Invest in your growth and shape your own path

The best investment you can make is in your own skills and growth. If you’re reading this, you’re already taking that seriously. My goal is to level up myself, my analysts, and you - by sharing practical, opinionated work on analytics: reporting, tooling, measurement, forecasting, examples/templates, and best practices.

Most companies offer professional development budgets. A subscription to this newsletter is typically reimbursable, and you can expense it through your company.

If you’re a current student or a recent graduate, I offer 50% off → click here. You can also email me at olga@berezovsky.me with your university/school name, and I’ll set you up.

This newsletter is my personal little project that isn’t affiliated with or sponsored by anything in any way, and it has grown where it is today because of your support. Thank you, thank you! ❤️

---

Hope you have a great end of the year ☃️. See you in January!