Building vs. Buying Your Growth Infrastructure
The trend of building custom tools instead of buying SaaS is real. But most companies get the tradeoff wrong. Here's the framework for deciding.
Writing
Thoughts on marketing infrastructure, paid acquisition, analytics, and the gap between growth strategy and data reality.
The Dashboard Your CEO Actually Needs
Your CEO dashboard is probably beautiful. It's also probably useless. It shows metrics that sound important but doesn't answer the actual questions your lead...
Stop A/B Testing Things That Don't Matter
Your A/B testing program exists to make your company feel scientific. Almost none of your tests will reach statistical significance, and that's by design.
Your Onboarding Funnel Is Leaking and Your Analytics Can't See It
Your analytics dashboard probably shows acquisition data in obsessive detail. But the moment users sign up, the visibility drops off a cliff.
How to Run an Incrementality Test Without a Data Science Team
You don't trust platform attribution. You shouldn't. But you can measure real channel impact with a simple test and no data science team.
Server-Side Tracking Isn't Optional Anymore
Client-side tracking is broken. iOS 14.5, cookie deprecation, and ad blockers killed the pixel. Server-side tracking is now the baseline.
The Real Cost of Broken Tracking
Broken tracking isn't an analytics problem. It's a budget allocation problem. A 20% measurement error costs you real money.
Why Your Google Ads Conversions Don't Match GA4 (And Which One Is Right)
You've noticed it: Google Ads says you got 487 conversions last month, but GA4 shows 392. Both platforms are counting the same website traffic, so why the di...
I Replaced a $40K/Year Attribution Tool With a BigQuery View
You don't need a $40K/year attribution tool. You need raw data that you control. Here's how to build reality-based metrics in BigQuery instead.
MER Is the Only Metric That Matters (And Nobody Wants to Hear It)
ROAS lies. CAC lies. Platform metrics lie. But Marketing Efficiency Ratio, total revenue divided by total marketing spend, tells you the truth. Most companie...
The Martech Stack Audit Nobody Wants to Do
Most companies run a bloated martech stack with redundant tools, broken integrations, and data conflicts. Here's how to audit it and what you'll discover.
Your Attribution Tool Is Lying to You. Here's How to Check.
Meta reports 1,000 conversions, but your CRM shows 650 closed deals. Most companies find their attribution tools overcount by 15-40%. Here's the reconciliati...
I Spent $40K Testing 127 Ad Variations Last Month (Here's What I Learned)
12 of them worked. Creative testing isn't about finding THE winner. It's about discovering why winners win, then making more things like that.
We Fired Our Best Performing Ad Because It Was Attracting the Wrong Customers
4.2x ROAS. Lowest CPA in the account. Scaling beautifully. And destroying the business. Here's how to tell when your best ad is actually your worst.
I Reverse-Engineered the Viral 'We Replaced Clay' Post. Here's What They Actually Built.
A lead gen agency claims they replaced Clay entirely with a $200/month Claude Code subscription and now process 272,000 leads per second. The post has thousa...
Most Companies Need a Practitioner, Not a Leader
Marketing has a missing middle -- a gap between junior marketers who are learning and CMOs who are leading. Here's why that gap is costing you.
The A/B Testing Market Is Solving the Wrong Problem. So I Built Cadence.
Most testing programs fail because nobody called the test. Because results got filed in Notion and never touched again. So I built something different.
Your Incrementality Test Just Proved Meta Doesn't Work (Here's Why You're Wrong)
You ran a geo holdout test, saw that Meta only drives 15% incremental lift instead of the 40% it claims, and turned it off. Bad move. Here's how to actually ...
Your LTV Calculation Is Off by 6 Months (And It's Killing Your Cash Flow)
Most LTV models calculate 'time until churn' but ignore payment timing. Here's why that gap between acquisition cost and actual payback is destroying your wo...
Why Every A/B Test Duration Calculator Lies to You (And What I Built Instead)
The standard A/B test duration calculator will tell you that with a 5% baseline conversion rate, expecting a 20% relative lift, you need about 18,000 visitor...
Built My Own Day Planner. The Code Was the Easy Part.
I canceled Reclaim.ai last month. Not because it’s bad. It’s actually quite good at what it does. But I found myself ignoring half of its features, which is ...
The Discovery Framework: How We Set a Baseline Before Optimizing Anything
You can't optimize what you can't measure. But the deeper problem is that most companies think they're measuring when they're actually guessing.
I've Been Rebuilding Analytics Setups. Here's What I Learned.
I’ve spent the last few months rebuilding analytics and tracking programs for three different companies. Each time, I walked into the same mess: event names ...
The Discovery Framework: How We Set a Baseline Before Optimizing Anything
You can't optimize what you can't measure. But the deeper problem is that most companies think they're measuring when they're actually guessing.
Build Chrome-Based Apps from Any Website to Stay Focused and Productive
Build Chrome-Based Apps from Any Website to Stay Focused and Productive