Synthetic control methods are a core technique for data scientists specializing in causal inference methods. One of the most popular packages to estimate synthetic controls is the CausalImpact package by Google.
In this real world data science causal inference problem, we look to see how well the Causal Impact synthetic controls could distinguish effects from noise.
Full video on YouTube: https://youtu.be/bVpctdxM3Cc
Here's a short tutorial and example of an Event Study, a popular and flexible causal inference model. Event study models can be used for a range of business problems including estimating:
⏺️ Excess stock price returns relative to the market and competitors
⏺️ The impact on KPIs across populations with staggered rollouts
⏺️ Impact estimates that change over time (e.g. rising then phasing out)
In my previous video, I introduced Google’s marketing mix model called Meridian which was just publicly released open source this past week.
In that video, I demonstrated how naive usage of prepackaged models can lead to terrible results. In this case, some minor changes led to dramatically improved — even impressive — results by Google Meridian.
I tested Google's new Marketing Mix Model aka Meridian with simulated data and it demonstrates how dangerous data science models can be in the wrong hands.
I simulated data so we know the exact relationship between marketing channels and bookings; we know precisely the share of bookings that each channel provided. But importantly, I tested the model using data that has real-world challenges including multicollinearity (spending can trend similarly across marketing channels) and endogeneity (marketers may choose to increase spending at times of high demand).
Large companies may have advertising budgets in the hundreds of millions of dollars so that even small efficiency improvements can be highly valuable.
Enter the marketing mix model. The MMM ingests a time series of advertising spend and impressions across channels and simultaneously estimates ROI across channels, providing opportunities for optimization.
Want to turn cold outreach into warm referrals, land research opportunities, and connect with industry professionals—all while building lasting relationships? Here’s how you can create a powerful, sustainable networking feedback loop:
Becoming a multimillionaire might sound like an impossible dream, but with a bit of consistency and some boring investments, it’s surprisingly achievable. Let me show you how.
Using a simple calculator, you can see the power of compounding and how even small monthly investments can grow into a fortune over time. For example
How are your interviews like real-world data science problems? 🤔
In predictive models, the most important feature is often whether the user has already done the action you’re trying to predict. Now put yourself in the shoes of the hiring manager
There's a trick to data science interviews: Across sectors, they are surprisingly similar. Companies like Amazon, Google, and Meta have shaped the process, so you’ll encounter common themes almost anywhere.
Best way to improve fast:
Donald Trump and Bernie Sanders recently aligned on a public policy proposal: capping credit card interest rates at 10%. But this rare bipartisan agreement may have unintended consequences.
Each of my last two jobs have come through LinkedIn, and I've had multiple offers that were initiated from LinkedIn messages. From my experience both sending and receiving messages, here are the sweet spots to prioritize:
Had an eye-opening conversation about 401k fees yesterday that I need to share. 🚨
A friend asked me to review his retirement portfolio, concerned about its performance. While I'm well-versed in the impact of investment fees, what I found was particularly frustrating - and it could be affecting your returns too.
© Copyright. All rights reserved.
We need your consent to load the translations
We use a third-party service to translate the website content that may collect data about your activity. Please review the details in the privacy policy and accept the service to view the translations.