Cheap Impressions, Expensive Attention: Where Your Video Budget Is Leaking

Where your media plan loses efficiency

Most video and programmatic campaigns are still bought and judged on CPM, reach and frequency. That makes sense: CPM is in every report and easy to compare. But it has a blind spot. 

A thousand impressions on Facebook and a thousand on YouTube look identical in a spreadsheet, even though their chances of being noticed can be very different.

When you split your budget by CPM, money flows to wherever impressions are cheapest, not necessarily to where people actually pay attention. One real campaign showed us how big that gap can be.

Case study: four channels, one yardstick

Our client, a Czech online retailer of premium cookware and kitchen equipment, ran a brand video campaign from October 2025 to mid-January 2026 with one goal: grow brand awareness. 

The video ran on YouTube, on open-web video inventory bought through DV360 (OLV), and on Facebook and Instagram. DV360 and YouTube alone reached over 2.4 million unique users.

For every channel we used Adelaide to measure how much attention the ads actually earned (the AU score), and converted the price into the cost of one unit of attention. 

This metric is called CPAU and the maths is simple: CPM ÷ average AU. The lower the CPAU, the more attention you get for your money.

CPAU = cost per attention unit. Source: Adelaide, brand video campaign, Oct 2025 – Jan 2026.

If CPM were the deciding metric, Facebook would win. Convert the price into attention and the ranking flips. 

On average, Facebook ads earned only a quarter of the attention YouTube ads did, so a unit of attention cost CZK 4.01 there. On YouTube, where impressions were 49% more expensive, it cost just CZK 1.58.

This was not a one-off. All four channels landed within 3% of Adelaide’s benchmark for their channel, so the gaps reflect how these channels generally perform in Adelaide’s data. If you buy a similar mix, there is a good chance they apply to your budget too.

How we measure attention, and why it holds up

Adelaide gives every impression an AU (Attention Unit) score. It expresses how likely an ad in a given placement is to capture attention and drive an outcome, such as brand recall. The model is trained on real campaign outcomes.

In 2025, its methodology was independently reviewed by the advisory firm MediaSense, which confirmed that higher AU reliably correlates with stronger awareness, recall and purchase intent. 

We compared several tools before choosing one. We went with Adelaide because it measures channels, formats and devices with one comparable metric, including YouTube and social platforms.

Same CPM, different attention: devices matter

Looking inside YouTube tells an even clearer story. CPM was practically the same on every device, at around CZK 113. Attention was not:

Source: Adelaide, YouTube part of the campaign.

For the same price, ads on a TV screen earned 24% more attention than on a phone. 

Through a CPM lens every device looks the same, so without attention measurement this difference would have stayed hidden.

The campaign also included non-skippable 15-second spots built for connected TV. Their CPM was higher than the campaign’s other YouTube ads (CZK 127.62), but with an AU of 71.97 their cost per unit of attention came to CZK 1.77. That is less than half the cost of attention on Facebook.

Why one number is not enough

If we went by CPAU alone, Instagram would be the first channel cut: it had the most expensive attention of all four (CZK 5.10). Yet for the same campaign, Meta estimated a higher ad recall lift on Instagram than on Facebook: 6.89% versus 4.24%.

The explanation is simple. According to Adelaide, AU deliberately does not rate the creative or the audience, only the quality of the placement. 

That is why we always pair attention with impact measurement. In this campaign, that meant a YouTube Brand Lift study comparing people who saw the ad with a control group:

• Brand awareness rose by 9.9 percentage points (control group: 8.2%).
• The lift was higher among women (+12.3 pts) than men (+8.7 pts).
• Non-skippable spots on connected TV showed a +14.8 pt lift after a manual calculation. The sample was too small to be statistically significant, so we treat it as a signal only.

Brand Lift and Adelaide point the same way: non-skippable TV spots performed well on attention and, based on early data, on awareness too. 

One number tells you what is cheap. Only a combination of measurements shows where your budget works hardest.

What it means for your budget

Imagine you have CZK 100,000 for a brand video campaign. At this campaign’s prices, here is what it would buy:

Calculated from this campaign’s CPM and CPAU, for a CZK 100,000 budget.

For the same money, Facebook gives you 50% more impressions than YouTube, but less than half the attention. If you split your budget by CPM, you buy volume that looks great in a report but carries less attention. 

That does not mean you should switch Facebook off. It means you should know what attention costs on each channel before you decide where the money goes.

5 questions to ask about your own media plan

1. Do you know what attention costs you, or only what impressions cost? If your report shows only CPM, reach and frequency, you cannot see how much of your advertising people actually notice.
2. Do you compare channels with the same yardstick? YouTube, social platforms and programmatic each report their own metrics. Without a shared metric, a fair comparison is hard.
3. Do you know how results differ by device? In our campaign, TV and mobile had the same CPM but a different cost of attention.
4. Do you measure brand impact? Attention shows placement quality. Brand Lift shows whether brand awareness actually moved.
5. Do you carry learnings from one campaign into the next? Attention data is most valuable when it shapes your next media plan.

If you answered “I don’t know” to even one of these, part of your budget may be working less hard than it could.

How we work with attention data

For us, attention measurement is not a one-off report. It is input for budget decisions. We follow the same process as in the campaign above:

1. We set up attention measurement across the channels you already buy.
2. We compare the cost of attention by channel, format, device and individual site.
3. We add impact measurement, such as Brand Lift, to see what attention does for brand awareness.
4. We propose a new budget split and test it in your next campaign, measured again.

Want to know what attention costs in your campaigns?

We will look at how you buy video and programmatic today and show you where your budget can work harder. We will suggest how to set up attention measurement, what it will tell you, and how to use it to split your spend across channels.