Paid search continues to be a critical channel for digital marketing, but understanding its real contribution to business growth requires more than looking at surface-level metrics like return on ad spend (ROAS). Incremental lift and marginal ROAS offer deeper insights into how effectively paid search campaigns drive new revenue versus capturing demand that would have happened anyway.

The Limits of Traditional ROAS Measurement

ROAS is widely used by marketers as a performance benchmark, indicating how much revenue is generated for every dollar spent on advertising. However, conventional ROAS figures typically represent attributed return rather than true incremental impact. This distinction is crucial because attributed conversions can include customers who would have converted via other channels such as organic or direct traffic without any paid search intervention.

For example, a brand keyword campaign may show an impressive ROAS because it captures high-intent customers searching for the brand. But many of these users might have come directly through organic searches or direct visits, meaning the paid ads simply shifted conversion attribution rather than producing new growth.

Case Study: eBay’s Brand PPC Experiment

An illustrative example comes from eBay’s controlled test of brand pay-per-click ads. They paused these ads for a subset of users to measure the actual lift. The outcome revealed that organic traffic compensated for much of the lost conversions, resulting in minimal revenue impact from turning off the ads. Despite the evident data, eBay reinstated the branded campaigns, highlighting common challenges in interpreting and acting on incremental lift data.

“Measuring incremental lift gives us a clearer picture of where advertising budget truly generates new business, rather than cannibalizing existing demand,” said a senior digital marketing analyst.

The Impact of Automation on Attribution

With increasing automation in platforms like Performance Max and Advantage+, marketers depend heavily on black-box systems that allocate budget across channels without transparent insight into causality. These systems excel at identifying easy conversion paths, often focusing on users who were already close to converting.

This leads paid ads to become the last touchpoint in a customer journey, inflating the perceived value of the advertising channel. Instead of acquiring new customers or expanding the market, automated campaigns risk spending on redundant touchpoints, thereby diminishing marginal returns.

Common Non-Incremental Signals Fueled by Automation

Automated paid search campaigns often emphasize:

Brand search campaigns targeting existing demand that would convert organically.

Retargeting users immediately before they complete a purchase.

Reporting metrics that overvalue ‘safe’ or familiar channels, masking inefficiencies.

Recognizing these patterns is essential for marketers aiming to optimize budgets and improve return.

How to Measure Incremental Lift and Marginal ROAS

Incremental lift quantifies the additional conversions or revenue generated solely by the presence of the paid campaign. Marginal ROAS calculates the return on spend for only these incremental conversions, offering a more accurate reflection of budget effectiveness.

To measure these, marketers can employ controlled experiments such as geo holdouts, holdback groups, or A/B tests where ads are selectively turned off to compare behavior. These approaches isolate the causal impact of campaigns beyond common attribution models that assign credit based on last or multi-touch attribution logic.

“Only with rigorous incrementality measurement can marketing teams confidently scale investments knowing they’re driving genuine incremental growth,” commented a performance marketing specialist.

Examples of Incrementality Testing Methods

1. Geo Holdouts: Pausing campaigns in designated regions to monitor revenue changes versus active areas.

2. Holdback Groups: Randomly excluding subsets of the target audience from seeing ads to observe differences in conversion rates.

3. Time-based Holdouts: Temporarily suspending campaigns during select periods to detect variation in performance.

Variable factors such as seasonality, competitor activity, and channel overlap should be accounted for to improve test validity.

Best Practices to Improve Paid Search Efficiency

Beyond measurement, understanding incremental lift encourages strategic shifts such as reducing spend on brand keywords that do not add net value, reallocating budget to prospecting new audiences, and refining retargeting windows to avoid wasting impression caps on near-converting users. Integrating lift data with lifetime value (LTV) analysis can further optimize prospecting efforts.

Additionally, using third-party attribution and analytics platforms can provide more transparent insights compared to platform-native reporting, mitigating bias and inflated credit claims.

As customers engage across multiple touchpoints, blending paid search insights with broader customer journey analytics will help create a comprehensive performance picture.

Conclusion

Relying solely on conventional ROAS metrics risks overestimating the effectiveness of paid search campaigns and misallocating advertising budgets. Embracing incrementality measurement techniques reveals true marginal gains generated by ads, enabling marketers to cut waste and make data-driven investments that drive actual business growth.

Paid search success today depends on going beyond attribution numbers to understanding causal impact, especially in an ecosystem increasingly dominated by automation and complex customer journeys.