A Media Operator is a media company for those building digital media companies.

Pivoting Brand Advertising Conversations With First-Party Data

As ad markets have weakened, there has been a significant pivot in ad spend from fluffier "brand campaigns" to more bottom of the funnel performance ones. CFOs are being tighter with budgets, so marketers need to have verifiable ROI on their campaigns. Brand campaigns tend to not fulfill that requirement.

What if there were a way to prove that brand advertising campaigns are more concretely tied to performance? Let's dig into that.

But first... A message about our sponsor, Omeda.

I spend a lot of time talking about first-party data for ad targeting, but there is an equally powerful type of targeting: contextual. Advertisers want to reach readers that are engaging with specific topics because it demonstrates intent. 

Omeda can help operators do that. 

On June 8th at 11am CST, Omeda is hosting a webinar where it’s going to dig into how B2B publisher, Questex, was able to drive nearly $6M in ad revenue through a unique product called Content Channel. This product allows advertisers to run highly contextual placements with full ownership of topics on the Questex website. 

By offering advertisers ad targeting both at the individual and contextual levels can unlock significant value. Register for the webinar to hear how Questex did it.

Now let's jump in...

***

According to a story in last week's Axios Media Trends (actual story unlinkable):

Analysts at New Street Research have slightly raised their 2023 forecasts for digital ad revenue growth in the U.S., thanks to resilience in performance marketing budgets that are mostly impacting search, retail and social media companies.

Yes, but: Demand for brand advertising "remains lackluster," per New Street, which is having an outsized impact on traditional advertising companies in television, print and radio.

Truly analog advertising is always going to struggle when times are tough. But all of us digital media operators are also feeling the pressure to perform. Brand awareness and thought leadership is being pushed aside for lead generation and affiliate deals.

But recessions are the worst time for advertisers to pull back on their brand building. According to research done by Analytic Partners:

How to Maintain Advertising Effectiveness in Challenging Times found that 60% of brands that increased their media investment during the last recession saw ROI improvements, according to analyses of hundreds of billions in marketing spend. Brands that increased paid advertising also saw a 17% rise in incremental sales, while those who slashed spend risked losing 15% of their business to competitors who boosted theirs.

...

Brand messaging outperforms performance messaging 80% of the time, so refocusing exclusively on performance messaging will lead to losses.

Part of the reason this is the case is that we, as humans, have short attention spans. We might be aware of a brand today, but as time goes on, that brand starts to go more stale in our minds.

I'm reading John F. Love's McDonald's: Behind The Arches, and one thing that is remarkable is how they were, at the time of the book's publication, the largest single product advertiser in the world. It made sense, though. You might eat a Big Mac today, but in a month, you'll have forgotten all about it. And so, you needed to be reminded again and again that McDonald's existed.

It's the same for all advertisers. In particular, this is true for b2b because there is so much focus on performance. Brand awareness is often forgotten because our budgets are smaller and we need to pump fresh leads into our sales team's CRM.

But this is a short sighted approach.

Colin Fleming, Senior VP of Global Brands, Events and Customer Marketing at Salesforce, told the B2B Institute at LinkedIn:

We found a great study on B2B buying behavior showing that two-thirds of the time, when a business decision-maker purchases software, they already have a brand in mind. And 94% of the time, the buyer ends up sticking with that brand. So if you’re not part of the original consideration set, there’s no way you’re getting bought.

That means 62% of the time (100*.66*.94), the buyer picks the brand they already know. Every B2B marketer who is only focused on lead generation is fighting for 38% of all the business. It's not quite scraps, but it's a minority of all purchasing behavior.

But what we also know is that not all brand awareness ad impressions are created equal. We need to show that the right people are seeing the ad. And this is where being able to leverage first-party data with brand advertising becomes so impactful. To do this, though, requires a CDP.

A CDP relates the data a user gives you about themselves with their behavioral activity. So, a user might tell you they are a Director of Marketing at Nike when they sign up for a newsletter. And then they read a story on your site about email marketing. For the most part, a good CDP can connect that reading behavior with the individual.

So if you can dig in and understand who is actually reading the content, that becomes powerful information you can share with your advertisers. Here's an example.

Let's say that you have an ESP that is sponsoring content on your site; it's a pure branding exercise. Historically, the data we've given advertisers is the number of ad impressions and the number of ad clicks. We would say something like, "well, since they're reading about email marketing, they must be interested in ESPs." And that's true to an extent.

But with a CDP, we can provide deeper analysis on who exactly saw the ad. And so, we could say that 35% of the viewers were Director+ at a retail company. Or, 10% were Directors of Marketing. Whatever data the advertiser cares about, we can zero in on that and tell the advertiser that the specific people saw the advertisements.

That means something. Once you can show them that, you can couple it with the statistic that 62% of all software is bought from the brand people know. Now you can start talking about how brand awareness campaigns can ultimately lead to a closed deal for your advertiser.

This couples nicely with other more performance-driven advertising. Let's assume the target buyer is a Director of Marketing. You can run an automated campaign using the CDP where every time a Director of Marketing reads a piece of content that the advertiser is on, they receive a dedicated blast. That might promote a piece of gated content. If the reader fills out that form, they are likely a warmer lead and more aware of who the advertiser is, which will typically result in a better outcome for the advertiser.

You can start to see how this gets very exciting across various brand awareness campaigns, such as editorial sponsorships and custom content. Being able to tell the advertiser exactly who saw the content (so long as you have the first-party data about them) gives you a major leg up to your competitors.

To do this, you've got to get your reporting right, though. It has to be easy to pull a report that shows who was consuming the content. But if you can, I think you'll be able to unlock some brand advertising even in this tough economy. Connecting it to a full funnel marketing campaign could be the secret to larger buys from marketers.

Whether they have a shot at winning any sort of suit against these generative AI platforms remains to be seen, but publishers are getting more vocal about their frustrations.

Marketing Brew reports that Digital Content Next, a trade association for publishers, is pushing back on these AI systems:

According to a draft of guidelines from Digital Content Next that was shared with Marketing Brew, “copyright laws protect content creators from the unlicensed use of their content” and “use of copyrighted works in AI systems are subject to analysis under copyright and fair use law.” Additionally, it says that “most of the use of publishers’ original content by AI systems for both training and output purposes would likely be found to go far beyond the scope of fair use as set forth in the Copyright Act and established case law.”

The document, which hasn’t been shared with members yet, also notes that “use of original content by [generative AI] systems for training, surfacing, or synthesizing is not authorized by most publishers’ terms and conditions, or contemplated by existing agreements.”

It's an interesting argument. We already know where Creative Commons comes down on these tools. In a blog post back in February:

In fact, the Supreme Court has recognized fair use’s importance in the development of new technologies, first in 1984, in Universal City Studios v. Sony and most recently in 2021 in Google v. Oracle. In Sony, the Court held that the Betamax videocassette recorder should not be sued out of existence even if it could potentially help people violate copyright law. Instead, because it held “substantial, non infringing uses”, the Court believed copyright law should not be used to stop it. Then in Google, the Court held that Google’s use of Google’s 11,500 lines of Java code was fair use, writing that the courts must consider fair use in the context of technological development.

Altogether, I believe that this type of use for learning purposes, even at scale by AI, constitutes fair use, and that there are ways outside of litigation that can offer authors other ways to control the use of their works in datasets.

Stephen Wolfson, Associate Director of Research and Copyright Services for the University of Georgia School of Law, has a number of good blog posts on Creative Commons about copyright and generative AI.

I'm no legal expert, so I'll take a different stance: I think it's too late. What are publishers going to do? Sue OpenAI into oblivion? For how much money? Will it try to get licensing deals? Again, for how much money? I explored that a couple of months ago in this piece.

According to OpenAI’s website, a 24,000 word prompt and response would cost $0.12 for API usage. Assume that your typical prompt and response is 750 words, you’re looking at 32 questions/answers. That’s $0.004 per question.

But let’s keep digging. According to HubSpot, “it’s estimated Google processes approximately 63,000 search queries every second, translating to 5.6 billion searches per day and approximately 2 trillion global searches per year.”

5.6 billion searches at $0.004 per prompt/answer is $21 million per day. Over a year, that’s approximately $7.6 billion. I recognize it’s not a perfect comparison since we’re talking about costs for API usage. But how much should publishers get of that $7.6 billion?

Suffice it to say, the path forward is unbelievably complicated for publishers. On one hand, I marvel at what can be done with this technology and think examples like Ask Skift are the way forward. On the other hand, I don't love giving up information for free. Being disintermediated doesn't feel right. But the systems simply don't exist to properly monetize these third-party platforms.

If this technology is going to take off, it's going to spread like wildfire. We are already seeing open source versions of ChatGPT. Once something goes open source, I don't see how you stop it.

I'll continue to follow what publishers do to fight back, but I see no path where licensing content to these AI tools makes up for the loss in revenue due to a significant drop in traffic. The best hope for publishers is that readers use these as exploratory tools, but then still seek out our content for more in-depth information. But it's still early to truly understand user behavior.

Thanks for reading today’s AMO. If you have thoughts, hit reply. Become an AMO Pro member to start receiving the Friday edition of AMO, for an invite to the AMO Slack, and for paid early bird registration to the upcoming AMO Summit on October 26th here in NYC. I'm in Portugal this week for the FIPP World Media Congress, so if you're there, let me know. Have a great week!