Where the attention was, and what it cost.
The report your client reads in ten minutes: how much of the attention was the media buy, the audience, the industry and the season, each measured apart, and what each looked-at thousand cost. Measured on real people, every number labelled.
Four numbers, each against the reference.
The first page of the example report, as the client reads it: “The display buy got attention on 28.4% of impressions against 27.0% for the ATTEX reference average: index 105.” Desktop did the work, at index 132 on 31% of the volume; mobile sat at the reference, index 101 on 69%.
Source: the ATTEX example report, § 01 Summary.
Four drivers, measured apart.
A campaign’s attention is one number until it is explained. The report takes it apart: how much came from the placements chosen in the plan, how much from the audience, the industry and the season, each measured on the same ad slots in the panel. So the next plan moves the lever that moved, not the budget.
In the example the media buy did the work; the industry and the season fall below the report’s 90% confidence bar and are printed as indications, never as results.
Index 105 is the campaign’s number. These four say why.
- The media buy01+5%attention+23%attention timemeasured
what is chosen in the plan
- The audience02−11%attention+13%attention timemeasured
men 25+, on the same ad slots
- The industry03+2.0%attentionno measurable effectattention timeindication, confidence 68%
consumer electronics brands
- The season04+6%attention+2.9%attention timeindication, confidence 85%
when the campaign ran
Source: the ATTEX example report, § 03 Drivers of attention. The media buy is the measured uplift of the placements themselves; the other three are measured separately on the same ad slots.
Each driver can be opened. The audience, men 25+, is split by age on the same ad slots: the youngest group looks less often than the rest of the panel, the oldest more, and all three look longer.
The older the man, the more often he looks, and the longer.
Index 100 = other respondents, mainly women. Gold: attention; cream: attention time.
Source: the ATTEX example report, § 03 Drivers of attention, the audience split.
Where the campaign was looked at.
Every publisher in the plan gets its attention rate against the ATTEX reference for the same screen, with its share of the volume beside it. In the example the volume sat with Expressen and Aftonbladet; the attention sat with Klart, GP and SvD.
Klart, GP and SvD well above the reference; hitta.se and Di at half of it.
Index 100 = the ATTEX reference average for desktop, 12.2%. Gold: at or above the reference; cream: below it. Share of campaign volume on a log scale.
Source: the ATTEX example report, § 02 Publishers, desktop chart and table.
Klart, GP and SvD well above the reference; hitta.se and Di at half of it.
Index 100 = the ATTEX reference average for desktop, 12.2%. Gold: at or above the reference; cream: below it. Share of campaign volume on a log scale.
Source: the ATTEX example report, § 02 Publishers, desktop chart and table.
Six of eight above the reference; Expressen and Di below it.
Hairline: the ATTEX reference average for mobile, 33.5%. In brackets: the publisher's share of the mobile volume.
Source: the ATTEX example report, § 02 Publishers, mobile chart and table.
The index turned into kronor.
Two prices per publisher: the cost per 1,000 impressions with attention and the cost per 1,000 attention seconds, both your CPM against what the panel measured in the same environment. In the example desktop had the higher index and the higher price: 243 SEK on mobile against 566 SEK on desktop.
A looked-at thousand cost least on Aftonbladet mobile.
Measured, volume-weighted: 304 SEK. * The whole display buy includes publishers predicted at the ATTEX reference level.
Source: the ATTEX example report, § 01 Summary, Attention per SEK.
A thousand seconds of attention, four times the price on Expressen desktop.
Measured, volume-weighted: 256 SEK. * The whole display buy includes publishers predicted at the ATTEX reference level.
Source: the ATTEX example report, § 01 Summary, Attention per SEK.
What to move, with the number.
The report ends with the moves for the next plan, as the example prints them. Calculated on the report's own levels: volume that is moved, not budget that is added.
- 01+6%mobile attention at Aftonbladet's level
Mobile: choose the slots on Expressen with care.
Expressen carries 20% of the mobile volume but sits below the average on mobile (27.1% against 33.5%) while desktop sits above it. Choose the slots on Expressen with care — at Aftonbladet's mobile level (37.6%), mobile attention would have been 6% higher.
- 02+3%impressions with attention on desktop
Desktop: lift the weakest.
hitta.se and Di sit at index 47 and 44 but carry only 2% of the desktop volume. Moved to DN, SvD and GP (index 156 to 166), that volume gives 3% more impressions with attention on desktop.
- 0382attention time index on desktop, mobile 146
Desktop: a shorter look.
When attention occurs on desktop it lasts 1.06 s against the reference's 1.29 s (index 82), while mobile sits at index 146. The ad slots are well chosen; review the desktop creative before the next campaign.
- 04243 SEKagainst 566 SEK on desktop
Budget: buy attention where it is cheap.
On mobile, 1,000 impressions with attention cost 243 SEK against 566 SEK on desktop.
Every number says how it knows.
- MeasuredAttention measured in the panel on the same publisher and ad format as the campaign.
- PredictedThe ATTEX reference level for the format times the report's factors for industry and audience. Marked with an asterisk in every table.
- IndicationA reference from a similar format, on thinner data.
No prediction sits under a measured heading, so the report goes in front of the client as it is. The example rests on 225,361 measured display impressions from 156 desktop and 75 mobile panelists in the campaign window.
Read one in full.
A Swedish display campaign, anonymised: brand, category, period, publishers, volumes and prices changed, the figures and the method as delivered.
A report for your own campaign, or a question about one, goes to Emelie Löfdahl, Co-founder and CEO: emelie@attexdata.ai. She replies within one business day.
Talk to Emelie