As DoubleVerify (DV) aims to make attention metrics a key part of its product mix, it tapped ad industry veteran Daniel Slotwiner to lead its Attention Lab initiative.
Although DV has been in the attention game for about three years now, the launch of its Attention Lab in October 2022 and the appointment of an SVP to head up the initiative represents a more serious commitment on DV’s part, Slotwiner told AdExchanger.
Slotwiner will spearhead the advancement of DV’s attention offerings. The Attention Index metric assigns attention scores to individual placements. Authentic Attention is a measurement and benchmarking platform and campaign planning tool that analyzes attention at the impression level.
In addition, the company will soon release a commercial product that will let advertisers activate campaigns using its Attention Index as a currency, Slotwiner said. For example, this information could become part of prebid segments so buyers can bid based on a predicted attention score.
Much like DV’s prebid targeting offerings, DV intends to collect a CPM fee from advertisers that use of its attention-based campaign activation product.
Measuring attention
Slotwiner last worked at Gopuff as VP of measurement, insights and product marketing management, and he previously conducted ad effectiveness research at Facebook and Google. He brings 20 years of media quality measurement experience to DV’s attention initiative.
DV’s greater focus on attention metrics is aligned with an industrywide push to develop more insightful methods for measuring ad effectiveness and new currencies for buying ad inventory. The current dominant currency – viewability – simply doesn’t go far enough for many marketers today.
But how does DV actually measure attention?
DV’s Attention Index assigns a numerical value that represents an ad placement’s relative attention score, using 100 as a baseline value. So, an Attention Index score of 25 would be considered on the low end, but 150 would be considered higher than average.
To calculate these scores, DV examines in real time more than 50 different data points derived from the ad’s appearance on the publisher’s page and how viewers engage with the ad. Ad-based signals include, for example, the size of the ad, the placement of the ad or whether a video ad plays with the volume turned on. User engagement signals encompass audio control interactions, clicks, screen orientation, how long the user’s cursor hovers over the ad and more.
One distinction between DV’s methodology and most other attention metrics is that DV does not use any eye-tracking data derived from volunteer panels or access to a device’s camera.
But it’s still early days for attention metrics, and attention-based offerings, including DV’s, need more development before they can become a currency in their own right, Slotwiner said.
Slotwiner has divided this task into three “buckets”: providing more transparency into DV’s attention-measurement methodology, marrying attention data to outcome data to prove that ads with high attention scores drive tangible results and expanding DV’s attention suite to other types of media inventory it doesn’t currently cover, like CTV and social media.
But what about publishers?
DV can access the various signals it uses to measure attention by virtue of existing media verification agreements between publishers and advertisers, Slotwiner said.
Because of these agreements, DV can place a tag within ad creative to measure brand safety and viewability of ads served on publisher sites.
Because no personally identifiable information or data tied to a specific user is being collected, DV does not need to obtain explicit user consent to measure signals based on how users engage with ads, Slotwiner said.
But the question of publisher consent is a thornier issue.
Some publishers have recently pushed back against the idea that their existing brand safety and viewability measurement agreements allow companies like DV to use any data obtained in the course of conducting these measurements in other commercial products.
For example, DV’s Custom Contextual offering uses impression-level data and page URLs to package publisher ad inventory into contextual segments that DV sells directly to advertisers, and some publishers believe their agreements with DV do not cover this use case. Using the same method to package ad inventory for sale based on attention score would likely raise the same concerns.
DV claims it does not use bots to crawl publisher pages for text-based signals like keywords and page-level metadata like ad dimensions.
When asked whether a commercial product based on attention metrics would be covered under existing brand safety and viewability measurement agreements or whether those agreements would have to be expanded upon, DV declined to comment at this time.
Update 4/20/23: A previous version of this article said DV uses bots to crawl for page-level signals from publisher sites as part of its attention score calculations. DV claims it only uses a tag placed within ad creative to examine ad placements and how end users interact with the ads, as well as the URL on which the ads appeared, rather than any page-level signals or signals about a user’s on-site behavior.
The previous version of this article also said DV’s upcoming product offering that will allow advertisers to purchase ads based on attention score will charge a service-level fee based on the CPM charged by the publisher for that ad placement. DV clarified that it will charge advertisers a CPM fee for ads purchased through this product, and that it has no insight into a publisher’s ad pricing.
The article also previously said that DV can collect data from publisher sites through its existing agreements with publishers. DV clarified that it can conduct its measurement operations on publisher sites because of measurement agreements between advertisers and publishers, rather than direct agreements between DV and publishers.
