Client Hub →
Theme
Glossary Attribution

Incrementality Testing

A testing methodology that isolates the incremental impact of a specific marketing campaign by comparing user behaviour with and without exposure to that campai

Also known as: incremental testing incrementality studies lift testing causal testing true lift measurement

What is Incrementality Testing?

Incrementality testing measures the true causal impact of a marketing campaign by comparing the behaviour of exposed users against a control group that didn't see the campaign. Rather than relying on last-click attribution, it answers the fundamental question: "What sales or conversions happened because of this campaign?"

The methodology typically involves randomly withholding a campaign from a statistically significant control group whilst running it normally for a test group. By comparing conversion rates, purchase behaviour, or other KPIs between the two groups, marketers isolate the genuine uplift attributable to the campaign alone – not confounding factors.

Why It Matters

UK media buyers increasingly face pressure to prove ROI in a privacy-first landscape where third-party cookies are disappearing and attribution becomes harder. Incrementality testing fills this gap by providing clean, causal evidence of campaign effectiveness independent of tracking limitations.

Without incrementality testing, marketing teams risk overstating campaign value. A user who converts after seeing an ad might have converted anyway – a phenomenon known as "cannibalisation." Budget allocated based on inflated attribution claims gets misallocated away from truly high-performing channels.

For FMCG brands, ecommerce players, and financial services – major sectors in the UK – incrementality testing reveals whether campaigns drive genuine incremental revenue or merely capture existing demand at higher cost.

When to Use Incrementality Testing

Incrementality testing works best for:

  • High-volume campaigns with sufficient budget and scale to create meaningful control groups (typically £50k+)
  • Direct response channels where conversion data is trackable (paid search, display, email, social)
  • Mature campaigns where you need precision ROI rather than vanity metrics
  • Major budget decisions where understanding true effectiveness justifies test costs
  • Privacy-first environments where first-party data and causal testing replace attribution models

Key Considerations

Duration: Most tests run 2–4 weeks minimum to capture enough conversions and account for purchase cycles.

Scale: Smaller campaigns may lack statistical power; control groups typically need 10,000–50,000+ users depending on conversion rate.

Cost: Testing reduces short-term performance for the control group, effectively costing the incremental revenue forgone during the test period.

Holdout fatigue: Running repeated tests on the same audience may reduce control group relevance over time.

Incrementality vs. Attribution

Attribution models (first-touch, last-touch, data-driven) assign credit to touchpoints within a journey. Incrementality testing asks a different question: does the campaign itself drive uplift? This makes it more reliable for ROI measurement, particularly for awareness and mid-funnel campaigns where attribution struggles.

Best Practice

Combine incrementality testing with attribution data for a complete picture: use attribution to optimise within a channel, use incrementality to validate whether that channel truly drives incremental business value.

Frequently Asked Questions

How much does incrementality testing cost?
Costs vary by platform and complexity, typically £5,000–£30,000+ per test depending on campaign scale and duration. The investment is justified when campaigns spend significantly (£100k+), as the cost to run the test becomes negligible relative to potential ROI insight.
Can you run incrementality tests on small budgets?
Not effectively. You need sufficient scale to create statistically valid control and test groups – usually a minimum £50k campaign budget. Below that threshold, sample sizes are too small to detect meaningful differences and test results become unreliable.
What's the difference between incrementality testing and A/B testing?
A/B testing compares two versions of the same thing (e.g., two ad creatives). Incrementality testing compares exposure vs. no exposure to measure true causal impact. Incrementality answers "does this work?"; A/B testing answers "which version works better?"
How long should an incrementality test run?
Typically 2–4 weeks minimum, though longer purchase cycles (luxury goods, financial services) may require 6–8 weeks. The test must run long enough to accumulate sufficient conversions and account for normal purchase delays.
Can incrementality testing work with iOS app campaigns?
Yes, but with limitations. IDFA deprecation and SKAdNetwork constraints make matching users harder. Incrementality testing works better on web or when you can track through first-party data, though some platforms offer privacy-safe incrementality solutions for apps.

Learn How to Apply This

Need expert help?

Our team can put this knowledge to work for your brand.

Request Callback