Incrementality

Attribution says which ad was clicked before a sale. Incrementality asks whether the sale needed the ad at all. A brand search that converts was often going to convert anyway.

Campaigns that look best on reported conversions, brand and remarketing, are usually the least incremental, so budget decisions based on them can move money away from what actually grows the business.

1. Google Ads does not show incrementality as a column.
2. Use experiments that split traffic, or Conversion Lift studies where available, to measure it.
3. Compare brand and non-brand results separately.
4. Look at new versus returning customers to see how much was repeat business.

A shop pauses its brand campaign for two weeks in one region and keeps it in another. Most brand sales in the paused region still arrive through organic results, so the brand campaign's real incremental value is far below what it reports.

• Treating every attributed conversion as caused by the ad.
• Running a holdout test too short or too small to give an answer.
• Pausing brand campaigns without checking whether competitors take the space.

The audit checks Oppy runs on this, each with its own page.

• A brand campaign letting Google decide what your own name is worth
• New customers against returning ones
• Performance Max with nothing stopping it buying your brand

• Experiment: A split test in Google Ads that runs a change on part of a campaign's traffic beside the original, so you can see whether the change worked rather than whether numbers moved afterwards.
• Brand campaign: A campaign that bids on your own business name.
• Attribution model: The rule that decides how credit for a conversion is shared between the clicks and ads that led to it.