TL;DR — Most “measurement” is a before/after that quietly credits your marketing for growth that would have happened anyway — seasonality, a rising baseline, a good quarter. This calculator subtracts a counterfactual (what the metric would have done untouched) so you see the real incremental effect of the play, not the apparent one. Enter your before, your after, and what your baseline did anyway. It tells you in 60 seconds how much of your “win” was actually yours.

The lift you measured probably isn’t the lift you caused

You ran a campaign. The metric went from 100 to 130. You report a 30% lift and everyone’s happy.

But here’s the problem: some of that 30% would have happened if you’d done nothing. The season turned. The category was growing. Last quarter was soft and this one was always going to bounce back. A naive before/after can’t tell the difference between the marketing worked and the tide came in — so it hands all the credit to the marketing.

That’s how budgets get poured into plays that did far less than the dashboard claimed, and how genuinely good plays get killed because a falling baseline ate their real effect.

Incrementality is the fix. It asks a sharper question: not “what happened after?” but “what happened because of the play that wouldn’t have happened otherwise?” The gap between those two numbers is the only number worth spending against.

This calculator does that subtraction in under a minute.

How it works

Enter three numbers: your average per period before the change, your average per period after, and what your baseline did anyway — a comparable channel or region you didn’t touch, or normal seasonal growth. The calculator builds a simple counterfactual, subtracts it, and returns your true incremental lift — plus how much of your apparent win was actually real.

Optionally, tell it how much this metric normally bounces on its own and it’ll flag results that are too small to tell apart from noise.

No login. No email required. The calculation runs in your browser.

Why “Measure” means a counterfactual, not a before/after

In the Growth Mapping methods, the Measure step has a specific definition: the effect of a play is the difference between what happened and what would have happened without it. That second half — the counterfactual — is the part naive measurement skips, and it’s the part that makes the number honest.

Davenport and Harris argued that companies winning on analytics ground decisions in what the data actually says rather than gut feel or convention.1 But a before/after feels like data while quietly smuggling in a convention: that nothing else was moving. It almost always was.

Thomke’s case for disciplined business experimentation lands on the same point — most “results” are confounded, and the only way to isolate cause is to hold a credible control against the treatment.2 You can’t manage what you can’t separate from noise.

The gold-standard version of this for time-series marketing data is CausalImpact — Brodersen and colleagues’ Bayesian structural time-series method, which builds a model of what the metric would have done from control series, then reports the difference with a credible interval rather than a single flattering point estimate.3 This calculator is the back-of-napkin cousin of that idea: one counterfactual, one subtraction, so you can see the gap. The platform does the rigorous version.

The logic is simple:

True incremental lift = (what happened) − (what the baseline would have delivered anyway)

Three inputs. One honest number. And it reframes every “the campaign worked” conversation you’ve ever had.

A guessed baseline is the back-of-napkin. A pre-registered control is the lever.

The calculator gives you the subtraction. But it’s only as good as the baseline number you type in — and right now that number is a guess.

The reason most teams can’t measure incrementality is that nobody decides what the control is before the campaign launches. By the time results are in, the baseline gets reverse-engineered to fit the story everyone wants to tell. That’s not measurement; that’s narration.

Hiilite’s platform pre-registers the control set at launch — the comparable channels, regions, or segments you’re not touching — and then computes the incremental effect automatically with CausalImpact (Bayesian structural time series), reporting posterior credible intervals instead of a single number. So when the Sense → Seize → Transform loop asks “did that play actually work, and should we run it again?”, the answer is a measured effect with uncertainty attached — proof, not a guess.

That’s the closed loop a before/after can’t build.

Read more: The Growth Mapping framework and Growth Mapping: the research behind the platform.

FAQ

What is incrementality (incremental lift)?

Incrementality is the portion of a result that your marketing actually caused — the lift that would not have happened if you’d done nothing. It’s the difference between what happened (your “after”) and the counterfactual (what the metric would have reached on its own from seasonality and baseline trend). A campaign can show a big before/after jump and have near-zero incrementality if the baseline was already climbing. Incremental lift is the only version of “it worked” worth spending against.

How do I measure whether my marketing actually worked?

Stop comparing before to after, and start comparing what happened to what would have happened. The cheap version: find a comparable channel, region, or segment you didn’t touch, see how much it grew on its own over the same window, and subtract that growth from your treated group’s growth — which is exactly what this calculator does. The rigorous version: a geo holdout, a randomized experiment, or a model like CausalImpact that reconstructs the counterfactual from control series and reports the effect with a credible interval. Either way, the principle is the same — isolate the play from everything else that was moving.

What’s a counterfactual in marketing?

A counterfactual is the answer to “what would have happened if we hadn’t run this?” You can’t observe it directly — you only get to run reality once — so you estimate it: from a control group, a comparable untreated region, normal seasonal patterns, or a statistical model trained on periods before the change. The whole discipline of measurement is building a credible counterfactual. Your true incremental effect is just your actual result minus that estimate.

Incrementality vs. attribution — what’s the difference?

Attribution divides existing credit among touchpoints — it answers “which channel gets this conversion?” and assumes the conversion was going to happen somewhere in your funnel. Incrementality asks the prior question: “would this conversion have happened at all without the marketing?” Attribution can make a channel look productive while contributing almost nothing incremental (think branded search clicks from people who’d have bought anyway). For budget decisions, incrementality is the truth and attribution is the bookkeeping — you want to know what to add, not just how to label what already arrived.

About the author

William Walczak is CEO of Hiilite Creative Group (2014–present) and a PhD candidate in Interdisciplinary Graduate Studies at UBC-Okanagan, where his doctoral research — Growth Mapping: A Mixed-Method Study of Growth Hacking — examines how small businesses can apply rigorous, data-grounded growth frameworks without a data team. He holds an MBA (UBC) and an Engineering degree (Simon Fraser University), and was named Marketing Strategy CEO of the Year 2023 (BC) by CEO Monthly.

His published research includes Walczak, W., Li, E. P. H., & Nelson, S. (2024), “Logarithm: A Cinematic Exploration of Time,” Journal of Customer Behaviour.

  1. Davenport, T. H., & Harris, J. G. (2006). “Competing on Analytics.” Harvard Business Review. https://hbr.org/2006/01/competing-on-analytics ↩

  2. Thomke, S. (2019). Experimentation Works: The Surprising Power of Business Experiments. Harvard Business Review Press. https://hbr.org/2020/03/building-a-culture-of-experimentation ↩

  3. Brodersen, K. H., Gallusser, F., Koehler, J., Remy, N., & Scott, S. L. (2015). “Inferring causal impact using Bayesian structural time-series models.” The Annals of Applied Statistics, 9(1), 247–274. https://doi.org/10.1214/14-AOAS788 ↩