Incrementality vs ROAS: making smarter B2C marketing budget decisions

In today's data-driven marketing landscape, the pressure to demonstrate clear returns on marketing investments has never been higher. CFOs want accountability, CMOs need to justify budgets, and marketing strategists must determine which metrics truly matter when allocating precious resources.

Two metrics commonly stand at the center of this discussion: Return on Ad Spend (ROAS) and Incrementality. While both measure effectiveness, they tell fundamentally different stories about your marketing performance.

Understanding ROAS vs incrementality

What is ROAS?

ROAS (Return on Ad Spend) measures revenue generated per euro spent on advertising. The calculation is straightforward:

ROAS = Revenue attributed to ads / Ad spend

For example, €10,000 spent generating €40,000 in attributed revenue equals a 4:1 ROAS or 400%.

However, traditional ROAS has a critical flaw: it relies on attribution models (typically last-click or platform-reported) that show correlation rather than causation. These models reveal which channel claimed credit rather than which channel actually caused incremental sales.

What is incrementality?

Incrementality answers a different question: "What would have happened if we hadn't spent on this channel?" It measures the true causal effect of your marketing activity on desired outcomes by isolating sales that wouldn't have occurred without your advertising efforts.

Incremental ROAS (iROAS) represents additional revenue generated solely due to advertising efforts divided by ad spend:

Incremental ROAS = (Revenue with ads - Revenue that would occur anyway) / Ad spend

Using our previous example, if that €10,000 spend generated €40,000 in attributed revenue, but €15,000 would have occurred organically, the incremental ROAS is only 2.5:1, not the inflated 4:1 shown by attributed ROAS.

Why the difference matters

The gap between attributed and incremental ROAS can be substantial:

• Facebook campaigns showing 4:1 attributed ROAS revealed only 1.8:1 incremental return in holdout tests
• Brand search campaigns reporting 6:1 ROAS in Google Ads delivered closer to 2:1 incrementally, as 60-80% of brand searchers would buy anyway. In other words by avoiding self-cannibalisation we could invest to other channels without anything happening with you bottom-line business results.
• Black Friday campaigns can show particularly stark differences, with last-click attribution often overstating performance by 4x compared to true incremental impact

This systematic overstatement of performance leads to poor budget allocation. Without measuring incrementality, organizations routinely:

1. Over-invest in channels that intercept existing demand
2. Under-invest in channels that create new demand
3. Misunderstand true marketing effectiveness

How to measure incrementality

There are two primary methodological approaches to measuring incrementality:

When to prioritize incrementality over ROAS

While incrementality provides a more accurate picture of true marketing impact, implementing full-scale incrementality measurement may not always be feasible or necessary. Here are scenarios when incrementality should take precedence:

Channel saturation assessment

When channels approach diminishing returns, incremental metrics become critical. For example:

• Paid search might show consistent 4:1 attributed ROAS regardless of budget
• Incrementality analysis might reveal ROI drops from 3.5:1 to 1.8:1 above €100,000 monthly spend

This insight enables you to reallocate budget from saturated to underfunded channels with higher marginal returns.

High cannibalisation risk

Incrementality becomes crucial when marketing activities might cannibalise organic traffic or other channels:

• Branded search campaigns often show high attributed ROAS but low incrementality (60-80% would convert organically)
• Promotions might appear successful but can cannibalize full-price sales (some retailers have found 12% cannibalization rates)
• Retargeting often claims credit for users who would purchase anyway

Cross-channel optimization

When trying to understand how channels work together, incrementality provides clarity:

• TV campaigns may show modest direct ROAS but create a halo effect that improves digital channel performance by 15-30%
• Podcast or radio advertising might improve paid search efficiency by 40% without receiving direct attribution
• Awareness channels may drive 3-5x higher customer lifetime value despite weaker initial ROAS

Budget reallocation decisions

Major budget shifts should be guided by incremental metrics:

• A channel reporting 5:1 attributed ROAS but delivering only 2:1 incrementally is less efficient than a channel reporting 3:1 attributed ROAS but delivering 2.8:1 incrementally
• Budget reallocation scenarios based on incremental metrics can produce substantial gains (e.g., shifting 20% from display to paid social predicted to increase incremental revenue by €340,000 quarterly)

Practical implementation of incrementality measurement

Building a hybrid measurement approach

Most sophisticated organizations use both metrics in parallel:

1. Track attributed ROAS for campaign optimization and tactical decisions
2. Use incremental ROI (from econometric modeling or holdout tests) for strategic budget allocation

This hybrid approach allows for both quick tactical optimizations and sound strategic decisions.

Technical implementation steps

1. Baseline establishment: Collect 18-36 months of cross-channel data (sales, spend, pricing, promotions, external factors)
2. MMM development: Build econometric models that account for baseline sales, channel effects, adstock (carryover), saturation, and external variables
3. Incrementality testing: Run geo-experiments or audience holdouts to validate model outputs
4. Refresh cycle: Update models quarterly, timed two weeks before budget planning cycles

Operational best practices

• Set model update triggers (e.g., if actual sales deviate from forecasts by more than 10% for two consecutive weeks or months)
• Implement a marketing investment board with cross-functional representation
• Translate complex model outputs into clear directives for media teams
• Start with a phased approach (e.g., four-month pilot focused on digital channels)

Case study: Balancing incrementality and ROAS

A European e-commerce retailer was heavily investing in paid social based on strong platform-reported ROAS of 5:1. However, a marketing mix modeling analysis revealed:

• True incremental ROAS from paid social was only 2:1
• Paid social was approaching saturation (diminishing returns)
• Display advertising showed lower attributed ROAS (3:1) but higher incremental performance (2.8:1)

The team developed a budget reallocation scenario:

• Reduced Facebook from €70,000 to €40,000 weekly (where marginal returns began dropping)
• Reallocated €30,000 to programmatic display
• Result: 18% increase in incremental revenue with no budget increase

This reallocation improved overall marketing effectiveness by focusing on incremental impact rather than attributed performance.

Common challenges and how to overcome them

Statistical significance requirements

Incrementality testing requires minimum traffic volumes for meaningful results. Solutions include:

• Longer test periods for smaller advertisers
• Meta-analysis across multiple small experiments
• Bayesian methods that incorporate prior knowledge

Organizational resistance

Finance and leadership teams accustomed to attributed ROAS may resist adopting incrementality. Strategies to build buy-in:

• Run parallel measurement for a transition period
• Educate stakeholders on the limitations of attribution
• Present case studies of successful incrementality-based decisions
• Frame the conversation around reducing waste rather than questioning past decisions

Implementation costs

Building robust incrementality measurement requires investment. To manage costs:

• Start with high-spend channels where impact will be greatest
• Consider managed services like Analytical Alley for faster implementation
• Start and continuously iterate to improve the insights quality.

The future of incrementality measurement

As third-party cookies disappear and privacy regulations tighten, incrementality measurement becomes even more valuable. Future developments include:

• AI-enhanced models: Machine learning improving prediction accuracy and automating scenario generation
• Unified measurement: Integrated platforms that combine MMM, MTA, and incrementality testing
• Automated optimization: Systems that dynamically reallocate budget based on changing incremental returns
• Privacy-first approach: Methods that work with aggregated data rather than relying on user-level tracking

Making the right choice for your business

The choice between incrementality and ROAS isn't binary. Both metrics serve important functions in a comprehensive measurement framework. However, strategic decisions about budget allocation should prioritize incrementality to avoid systematic misinvestment.

By understanding the true causal impact of your marketing spend, you can eliminate waste, optimize channel mix, and demonstrate genuine business value to stakeholders. Start by implementing basic incrementality testing in your highest-spend channels, then gradually build a more sophisticated measurement approach that balances tactical attribution with strategic incrementality.

When making the transition to incrementality-based decision making, remember that the goal isn't perfect measurement but progressively better decisions. Organizations that adopt this approach can reduce wasted ad spend by up to 40% through disciplined measurement and testing.