Three distinct studies. Web attribution matches devices to your site. Search lift reads search interest in your markets. Halo effect measures whether your other paid channels performed better because OOH was running.

Exposed device to site conversion, one to one, inside the attribution window. It is the only one of the three that requires the units to carry attribution data, and the only one that requires a pixel.

A comparison of branded search volume in your campaign markets against control markets and against competitor terms, over the period before, during, and after the flight. It is built from search volume data, not from your accounts.

You supply your brand terms and the competitor terms you want indexed against. Nothing else, and it can be run after the campaign. One feasibility check applies: your brand term must have had search activity in the past two months. A term nobody has searched recently has no baseline to move against.

It measures search volume, not causality. A rise in the chart is a signal worth having rather than proof that your campaign produced it. If what you need is a demonstration of cause, this is not the study to reach for.

Whether your other paid channels performed better in the OOH markets than the model predicted they would have alone. It reads impressions, clicks, and conversions from your connected accounts.

It measures engagement with your search ads, not organic keyword search volume. That distinction confuses almost everyone, because both sound like search. Organic search interest is search lift; paid channel performance is halo.

Authorized view or read access to Google Ads, Google Analytics, and Facebook. You are not granting the ability to change anything. If authentication is a blocker, a manual reporting export can substitute.

Twelve months of pre-period data in both the test and the control markets is preferred, across whichever of those accounts you connect. It is a preference rather than a hard gate, and a shorter history produces a weaker read rather than no read.

Gaps in the connected data weaken the result rather than breaking it, so a partial authorization produces a weaker read, not an error message.

Halo needs more than one market. A single-market campaign has no control and no usable halo read. Search lift still works in that situation.

One point of overlap is worth naming, since it is the source of most of the confusion: the Google Ads connection informs both analyses. That does not make them one study. Search lift reads organic search volume in the market; halo reads performance inside your paid accounts.

Google Ads, Google Analytics, and Facebook are the connections available today, with one Facebook ad account per campaign. There is no TikTok integration and no Taboola integration; both have been asked for.

Where no integration exists, the same causal analysis can sometimes still be run on an exported data file, so the absence of a connector is not always the end of the conversation. A large retail advertiser has been handled this way.

Myth: Halo effect measures increased keyword search volume.

Reality: It measures increased engagement with your search ads. Keyword volume is search lift, a separate analysis.

Myth: Web attribution, search lift, and halo are three views of one thing.

Reality: They are three separate studies with different inputs, different outputs, and different requirements.

Myth: Halo requires having run paid social both with and without OOH.

Reality: It does not. For a digital OOH only campaign, lower funnel attribution is the path.