In this post:
- The changing definition of “Transparency” in ad tech
- Misrepresentation of platform transparency is a common problem
- How CTV breaks some platforms’ claims of transparency
- Why lack of total transparency has implications for campaign measurement and management
In this industry, transparency has become a cause or a pledge that is not unlike the environment. Everyone is for it, at least on the surface. But most accept the numbers they see within a DSP without asking the tough question of how those numbers came to be.
While it wasn’t uncommon years ago for ad networks or exchanges to hide behind sophisticated models, panels or proprietary relationships as a way to not disclose information about who they work with, today that position is no longer acceptable.
Rather, with the advent of low-cost cloud-based computing and data lakes, it’s seen as table stakes for brands to get information on seemingly basic requests such as: “where are my ads running?” or “how many conversions did my media campaign elicit?”
Everyone’s transparent, or so they say.
But more and more we find the definition of transparency stretched well beyond its core meaning.
Is the ad industry being transparent about transparency?
Take the two seemingly simple but vitally important topics of managing a campaign’s reach and frequency and measuring conversions. Anyone who consumes media, particularly CTV, knows this is a challenge (even if they don’t use the terms “reach” and “frequency”).
But chances are when brands work with various vendors they hear that managing reach and frequency isn’t a problem. The conversation surrounding this subject is often one-sided, something along the lines of, “Hey CMO, here’s how you’re performing.” Spoiler alert: The numbers look good.
The issue of consumers being bombarded with the same ads, again and again, persists, if it’s not getting worse. This, on top of marketers’ acceptance of lower-than-expected conversions or “modeled” conversions that just don’t seem quite right, only make real transparency that much more pressing.
Transparency matters for campaign management and measurement
A big part of the problem is that so many industry players claim to be able to reconcile targeting and measurement across devices, but never quite say how.
As an example, consider CTV. For starters, there are basic limitations in CTV ad targeting that are too often ignored by many vendors on the market. Historically, if you think about programmatic advertising at large, there have been four types of consumer identifiers available to brands for targeting:
- Mobile ad IDs (MAIDs)
- Company identifiers (such as Unified ID)
- IP addresses
Cookies, MAIDs and company identifiers are non-existent across CTV, and there are 1.5 billion IP addresses flowing through the programmatic ecosystem today. And this is not a new reality, although likely a surprise to DSPs built on the backbone of pre-CTV, cookie-and-MAID-based offerings. So how are reach and frequency, along with proper conversion attribution, being managed, exactly?
Many media companies are looking to build out their one proprietary, cross-platform identifier, and they claim to be able to manage frequency within their own universes. And there are numerous initiatives underway to create some sort of interoperable TV identity system.
But while we’re waiting for all this to be solved or consolidated, brands should start by at least asking their demand-side platforms some pressing questions. How are they managing the transition from cookies to other IDs? How are they connecting the dots across devices exactly? Where can I see my log-level data?
Transparency’s impact on reporting and optimization
And the biggest question: How are you tracking conversions across all these channels? Because as you know, most of us don’t buy things via the TV remote (yet, anyway). Across the Adelphic software, representing campaigns in the tens of thousands, 75% of conversions happen on a different device from where the ad occurred.
If I’m a marketer, I have a hard enough time capping reach at three times on every different platform and device. How can I know what actually worked without the right connective tissue? What does capping reach at three times even mean? The answer depends on the DSP’s technology. As marketers are increasing the variety of channels and tactics they’re buying across, capping becomes that much more difficult.
Cookie- and MAID-based DSPs will find it difficult, as 60%+ of inventory no longer carries cookie ID. Mobile phones only have MAIDs in-app and some cookies in web (depending on the browser). The cookie on your in-web mobile phone will be different from each of your laptops, desktops and tablets. And, of course, CTVs don’t contain either cookies or MAIDs. While your DSP may claim to cap at three, the three is per identifier they see flowing through the programmatic bid stream. Hence, transparency is a must.
The Adelphic by Viant approach to transparency in advertising technology
At Viant, we are able to map cross-channel campaigns using a household ID. And in terms of delivery, we share log-level data.
Clearly, this is not the only methodology out there. But no matter what, it all comes down to the idea that brands need a consistent identity, both for targeting and measurement.
Otherwise, just how transparent can you really be?