What is the halo effect in marketing?
What does the halo effect look like?
Someone sees a video ad. They do not click. A week later they search your brand name and buy. The search ad or organic result gets the credit. The video gets nothing in its own report.
That is the halo: the video did something, but the evidence sits elsewhere.
A measurement vendor, Sellforte, writes that upper-funnel video and reach campaigns are especially likely to create effects outside the path the platform can observe1. So their platform ROAS can look weak. That is a vendor's view, and it points in a direction without giving a size.
Why is it hard to measure?
Each platform only sees its own touchpoints. A click-based report cannot see the person who watched and did not click. So the effect, if there is one, appears in a different channel's numbers.
The same logic works in reverse. Branded search often gets credit for sales that would have come anyway. That is cannibalisation: paid activity taking credit for sales that would have arrived by another route.
To see a halo you need a comparison. Switch the campaign off in some places and watch what happens elsewhere. Methods such as MMM vs MTA vs incrementality tests differ in how well they can do this. Marketing-mix modelling uses aggregated data without user-level tracking, so it can see across channels.
Our view: A halo is a hypothesis. It earns budget only when something would measurably drop without the campaign.
Easily confused with
Halo effect vs Channels adding up to more together
A claim that the combination is worth more than the parts. Halo is narrower: one channel's effect on another.
Halo effect vs Spillover
In a test, ads from a test region reaching people in a control region, or the other way round. A measurement flaw, not a channel effect.
Halo effect vs Cannibalisation
The reverse problem: credit taken for sales that would have arrived anyway.
Buzzword alert
"The campaign is driving a halo effect"
Usually means: the campaign's own numbers are weak and a benefit is assumed elsewhere.
Ask: was there a group or region that did not see the campaign?
"Brand search is up because of our video"
Usually means: two lines moved in the same period.
Ask: did brand search also rise in weeks or places without the video?
FAQ
Is the halo effect real?
Yes, it can be. People are influenced by ads they never click. The problem is claiming it without a measure.
How do you measure a halo effect?
By comparison. Pause or reduce the campaign in some places and watch results in other channels. Modelling across channels can also help.
Is the halo effect the same as spillover?
No. Halo is one channel affecting another. Spillover is a flaw in a test, where the groups are not cleanly separated.
Let’s find what’s actually working.
Next step. See what your ROAS figure leaves out. Use the Is your ROAS real? calculator.