Find out what your marketing actually causes.

Why incrementality testing matters

Every ad platform reports its own results, and every one of them over-credits itself. Incrementality testing answers a simpler question: what would have happened without this spend? By holding out a group of people or markets and comparing them to a group that saw the ads, you measure the lift a channel truly causes, not just the conversions it touched.

Brands usually bring me in when they want to:

• Prove whether a big channel, like paid social, search or TV, is actually driving new revenue
• Validate or calibrate a marketing mix model
• Measure channels that platform attribution can't see, like CTV, audio, OOH and brand campaigns
• Build a regular testing program instead of one-off tests

What I've built

At Credit Karma, I ran the paid marketing experimentation program behind a $300M+ budget, with geo tests across Google Search, Meta, TikTok, Apple Search Ads, DV360, Moloco, linear TV and CTV. I automated geo-test monitoring in Python and SQL, replacing six-week readouts with weekly checkpoints so the team could run more tests at once. Those results calibrated the marketing mix model that guided spend as budget scaled from $100M to $300M.

At Talkiatry, I built the lifecycle experimentation roadmap and revenue-impact reporting, taking email and push testing from zero to 10 tests a month. At Robinhood, I use geo-lift experiments and brand lift studies to validate upper-funnel brand spend.

How I design incrementality tests

• Geo experiments and geo-lift. Turn spend up, down or off in matched markets. Works for nearly any channel, including offline, and doesn't depend on user-level tracking.
• Holdout and conversion lift tests. Use platform tools or audience holdouts where they're reliable, with an independent read on the results.
• Synthetic controls and diff-in-diff. When a clean randomized test isn't possible, these build a credible comparison from historical data.
• Power and duration planning. Every test is sized before it launches, so you know it can detect the effect you care about and you don't burn budget on inconclusive results.
• Readouts leadership can use. Each test ends with a clear call on what to scale, cut or test next.

Engagements

I work as an independent incrementality testing consultant on a project or fractional basis, from designing and reading a single high-stakes test to building an ongoing experimentation program and roadmap. I do the design and analysis myself, so you work directly with me.