Beyond the Click: A Trust-Based Model for AdTech

The “Web” we live with intimately is continuously evolving.

The “Web” we live in is less and less a trusted place for “Judy Consumer” because too much is unverified and unverifiable. The Web’s trust gaps affect each and every one of us:

• We can’t know for sure whether the person reaching out to us is real, a troll, a bot or worse.
• We have no tools to spot disinformation deliberately meant to deceive.
• We can’t easily figure out which content is real, parody or designed to defraud.  
• There are no easy ways to know if an eCommerce store is reliable, trustworthy or a scam.
• We are an easy mark when companies deliberately cloak their identities.  
• We can’t trust the reviews we see for a product are authentic.
• We are subject to many online fraud scams such stock or bitcoin hype or those insipient “get rich quick” financial newsletters that promise life changing “tips.”

In short, for most people, online activities make it feel like we are in a shooting gallery being targeted by an army of untrusted entities. We are, in every sense, sitting ducks.  

To unpack how we got here, let’s go back to how it started, beginning with the basic terms like Internet or Web, that we often and mistakenly use interchangeably:    

• The Internet is a global network of computers (hardware) that are connected through telecommunications and optical networking. The internet is like the digital road system used to transport “goods” like data and content.
• The Web is the software that allows users to access, edit, discover, and share information linked by hyperlinks and URLs. Think of the Web as the trucks on the road carrying the goods, i.e. content you see in your browser when you’re online.

While both the Internet and the Web are important, the Web has gotten the most attention because that is what people experience so that is where we will turn our attention to unravel the mystery of the Web’s trust gaps.

The evolution of The Web.
It’s easy to forget that the Web is only about 30 years given its outsized presence in all our lives. The Web started in the early 1990s, often referred to as Web 1.0 and it emerged as a content serving engine powered by telecommunications hardware. During this period, The Web was a one-way street where users could read content but not much else. Some examples include Yahoo! and Geocities.

Then, about 10 years later (2000’s), we got Web 2.0 where users could create and share content. This is the moment when u-gen content reigned supreme with blogs and online forums exploding. Noteworthy, Facebook, Twitter, and YouTube got their start during this period because “everyone” could be a publisher looking to reach as many people as possible.  

Now, we live in the era of Web 3.0 – sometimes called the Semantic Web. The term “Web 3” coined in 2014 by Ethereum co-founder Gavin Wood, suggests the decentralized future of the Web.  In this schema, there is not a single, centralized system but instead, a myriad of applications that are managed “at the edge” – with users. Some dominant applications of Web 3.0 are blockchain, AI and virtual reality.

Pulling back the lens even further, there are some key observations we can make about the Web throughout its evolution:  

• The terms 1.0, 2.0 and even Web 3.0 follow the nomenclature structure of software releases which seems mismatched for the huge human role the Web plays in our very human lives.  
• The trend to push more and more controls to the “edge” – end users – creates a different set of challenges that invites chaos. Deep fakes, online scams and loss of trust in online interactions gives us a glimpse of the dangers ahead. Now, with AI, the risk profile exponentially increased. Without guardrails, pushing technology to the edge can release a torrent of hurt far more pervasively than ever before.
• The Web 1.0, 2.0 and 3.0 gave scant attention to the verification/ trust technologies in ecosystem. While the Internet has some verification technologies available like SSL, when it comes to the Web, there is virtually nothing. Identity verification can be gamed. Content sources cannot be verified. Privacy platforms do not ensure user privacy.  

Why was “trust” MIA in every version of the Web?
You are likely wondering why the Web was left devoid of a technological trust layer when trust is a central feature in our physical world especially when in the real world:

We know how to verify money – mostly.

We know how to vet people – mostly.

We can calibrate the risks we are willing to take in transactions. For instance, if we buy a $10 product in a store and it is a dud – our risk is well understood and very limited.

The Web is entirely different because there is a much higher risk profile in every respect. A bad actor can wreak far more damage to our online identities than can be easily assessed upfront. Despite this, trust tech is MIA on the Web.

Why?

The answer, it turns out, is hiding in plain sight once you follow the money.

The killer application of the Web was digital advertising driven by technology, a.k.a. adtech. Despite the variety of adtech players, there was one business model to adtech which was some combination of “scale” media buys (powered by programmatic) and surveillance of “profiles” across the Web.

There is a lot of money to be made with this single business model – track everyone (surveillance) and buys media everywhere and anywhere (scale). Technically, the “surveillance” and “scale” business model is relatively straightforward to execute. The issue is that in this tonnage business model, traffic and profiles can be faked or gamed which left a huge trust gap allowing adtech players to generate huge profits.

Advertisers started to push back on the lack of any verification technology, so a few companies launched traffic verification platforms (like IAS and DoubleVerify) as a trust layer. This was largely done for optics since it turns out these verification platforms were in on the scale game too. Traffic verification platforms have a huge conflict of interest because these firms made money by scoring ALL traffic again and again and again. These verification companies had no financial incentive to stem the flow of bad traffic because more traffic equals more profits, (this article goes into detail – https://trustwebtimes.com/traffic-authentication-the-most-nettlesome-issue-in-ad-tech/).

This is why almost no verification is built into the system, because the less verification, the better it was for adtech’s bottom line:

• Adtech can sell cookies – real or fake.
• Adtech can sell “engagement” – real or fake.
• Adtech can sell ad placement on fake publications that are designed just to capture advertising dollars, called Made For Advertising (MFA) sites.

A knock-on effect of the scale/ surveillance model was that media companies were given a green light to “amp up” their traffic numbers with fake traffic to generate more advertiser revenue. With traffic verification largely ineffective, fake traffic on publisher sites skyrocketed so that now no one even agrees about the scale of the fake traffic problem. Estimates range from 30% – 90% (pretty broad range). Even at its lowest level of 30%, 1 in 3 “viewers” of ads are not real. Imagine, for example, we ordered 1,000 widgets but only got 700. No one would accept this ordinarily but when it comes to adtech, there is a general shrug of acceptance because “that’s the way it is.”

Adtech’s monolithic business model of detachment is unsustainable.
Now we know why trust tech is nowhere to be found in today’s adtech. There is a lot of money to be made with adtech’s single business model – track everyone (surveillance) and buy media everywhere and anywhere (scale). If some of the traffic is fake – oh well – that is just the price to pay to reach a large audience on the cheap.  If some web pages are brand unsafe – oh well – that is the price of programmatic’s muscly efficiency.

It has taken over 15 years for advertisers to realize adtech’s downsides in terms of waste and user privacy issues are not sustainable. A reasonable adtech response would have been the introduction of trust technologies. It never happened because that would gotten in the way of profits.

In fact, adtech doubled down on its faux reality environment and this is where our story takes a very dark turn indeed. 

The pre-digital ad days, there was money to be made by publishers and advertisers because there was synchronization of verification technologies.

• The circulation of magazines was audited by third party companies.
• All TV commercials had to go through verification of claims.
• Ad placement in magazines could be verified via “tearsheets” – actual proof the ad ran in magazines
• TV audiences were verified by large research companies

As a result, all this verification created an economy of scarcity:

• Scarcity of quality media outlets with quality content worthy of brand support
• Scarcity of real audiences for brands that could convert
• Scarcity of real engagement metrics from downloads to clicks

Scarcity created value that quality publishers and local news could monetize profitably. Real audiences could rely on trusted publications to inform them. Advertisers could then reach real people who are in a position to convert. Consumers win. Advertisers win. Publishers win.

Quite deliberately, adtech threw out the business model of verification and scarcity, replacing it with a business model that valued tonnage – of profiles tracked/ delivered and media bought.

Taking this model to its logical conclusion, adtech deliberately evolved to be more profitable the more detached it became from reality.

It is hard to “scale up” high-quality real audiences and high-quality content to support the tonnage model adtech relied on. Enter fake profiles and blind media buys. Adtech eschewing verification is not a glitch but a feature. Adtech didn’t have to bother with niggling issues of verification of audiences and content as long adtech sell could “people” and media” with cool sales pitches and seductive “tech” stories they would tell at industry events.   

The human costs of an untrusted Web. 
In 2024, it is obvious the Web is creaking under the weight of its untrustworthiness. The downstream human effects of an untrustworthy Web were catastrophic:

• Adtech powered the distribution misinformation (incorrect information not intended to deceive) and disinformation (deliberate untruthful information) easy and at scale. This opened the door for political interference and conspiracy theorists to have a platform unavailable before.
• Adtech could amp-up audience and engagement numbers which led to a flood of fake accounts and trolls designed to sow cultural discord in the U.S.
• Adtech made it easy for bad actors to advertiser cheaply and at scale via programmatic platforms which allowed a wide swath of dishonorable eCommerce sites to thrive.
• Adtech encouraged the creation of sites just designed to get ad dollars without a shred of editorial or journalistic integrity. “They’re eating the cats” is a poster child of this issue that makes it hard for individuals to get trustworthy news.

The list is much longer but what is clear is that trust needs to be the defining feature of the next Web. If so, what does that really look like and how do we do it?

Welcome to The Trust Web – The Alternate Web Business Model.
At the 100,000-foot level, we understand adtech’s business model is designed to exclude trust technologies given all the financial incentives in place to keep trust-building verification out.

In effect, adtech became a massive “tonnage push engine” built to be profitable and by design to be wasteful too.

With only one business model in the game, there were no alternate business model that could keep in check adtech’s worst impulses. The place to start, therefore, is to be inspired to create an alternate business that is a synthesis of the best of analog ad dynamics with the best that adtech offers marketers. It is in this intersection a different model can emerge that reverse the push dynamics of today’s adtech into a powerful “outcome-based pull engine.”

Data for an alternate adtech business model.
The key, then, to re-imagine adtech as a pull engine starts with data – the linchpin of every adtech system. However, the data in this business model is completely different than the surveillance data of adtech today.

This new data set pinpoints topics or themes audiences need see to convert. More important, this topic data is predictive so brands can focus resources on highest performing topics.

Waste is reduced. Guessing about outcomes is reduced. Tracking people is not needed.

This topic approach harkens back to pre-digital days when topic targeting delivered high performing results for advertisers because people chose paid subscriptions in magazines that matched their interest. In this “pull” system, advertisers bought topic-centric media such as luxury marketers who ran ads in Sailboat Magazine or car companies who ran ads in Car & Driver. Topic targeting was the central targeting mechanism in pre-digital adtech and it worked incredibly well.

Some might be thinking, topic data does not sound muscly enough to get the job done. This is true for marketers that need tonnage for ubiquitous products from pens to cola. However, most marketers need quality tonnage to convert audiences and topics is the most efficient way to get there. Topics can drive this alternate business model because it is a high performing alternative to the adtech’s dominant scale/ surveillance business model.  

• Topic data captures better intent signal from real audiences. This will allow for more accurate outcomes.
• Topics represent a consumer friendly “pull” model of adtech as in, “if you build it they will com” especially if the content is highly desired and desirable.  
• Topic data is what powers the conversion journey – all without the toxic need to track anyone – anytime.

If this alternate is so effective, one should wonder, why haven’t other adtech firms latched onto the topic bandwagon?

The answer is plain enough. To truly take this concept of topic/ theme data into high octane performance data turns out to be a technical high bar to clear. Topics are complex concepts that are deeper than simple keywords but more precise than interest classification. Topic data to be groundbreaking requires a deeply powerful semantic language model that understands digital text in natural human language. While it is technically easy to spot a keyword on a page, it is far more difficult to translate web content into topic themes that are resonance in human terms.   

Even Google focused on keywords in its main ad products and are only now pivoting toward search terms. Their misfires on that contextual path have been widely ridiculed and reported.

This explains why so few tech companies actually invested in topic/ theme data even though the outcomes and rewards are significant. It is really really much harder to do than the current data focused on scale and surveillance.

Can topic data really compete with adtech tonnage/ push business model?
Simply put, yes because a topic-centric business model is an outcome centric performance engine for brands.  This type of data can deliver better outcomes with less waste and cost for advertisers. This changes everything.   

• Surveillance data of today’s adtech versus topic data. Topic data (like Topic Intelligence) dismantles the need to track people. This simplifies privacy compliance challenges and instead, informs a brand about what content is needed for the topic journey to conversion. With contextual topic data, a brand knows what topics are worth investments. No need to amass huge profile data stores and spend money on reconciling digital IDs. GDPR compliance worries are in the rear-view mirror because people are not tracked – only topics relevant to a conversion journey. Most important, topic data is highly transparent and verifiable with no financial incentive to target a lot of “profiles.” 

• Scale media buying versus topic-based media placements  – By changing the data foundation of adtech to a topic paradigm, we can dismantle the scale media buying cylinder too. Well-defined, brand-specific topics will, naturally, limit the number of places a brand should run advertising. In this model, there are never going to be an endless number of pages available to run ads because it is topic specific to the brand. This type of contextual targeting does not “do” scale media buying but instead relies on content centric ad buys like direct buying or sponsored content ads. It is more expensive in labor to execute these programs but since there is not as much “scale waste,” the extra labor costs is covered by the reduced “scale” media buys. Another major plus to a topic schema is that brands don’t have to be adtech experts in attribution. They can measure the effects of topic campaign and verify the sales results they get. It is a highly streamlined attribution approach.

The alternative that works because it is brand outcome based.
Taken together, topic data offers an alternative to surveillance/ scale adtech business model because it naturally right-sizes media and profile scale. This model re-introduces the laws of scarcity to reflect the real world where real sales transactions can occur.

Most important, this business model works because it works in the real world. Advertisers get real attention from real people because well-chosen topics allow audiences to self-select. This high-intent audience engagement and conversion is the key to this trust-based business model. The entire business model now shifts in favor of the consumers, publishers and advertisers because:   

• Advertisers can trust where their media money is going (topic-based in human-scaled media) and who is being targeted are real people interested in brand-specific topic. They can trust the results they see.
  
• Publishers can trust that Web will help them monetize their content ethically because they can charge premium CPMs given the high quality of content based on high performing topics.
  
• Consumers can trust their privacy is being respected because in this system, people are not tracked. In this system, ads are contextually relevant and thus a welcome part of a user’s experience. 

Topic data is an innovation that leaps frogs over the performance of keywords. Nor is it just an iteration of current contextual tech. Once trust becomes the central organizing model through a new data class, this creates new dynamics for adtech itself.   

No one business model fits every marketer but by diversifying digital marketing’s business model, we will see a healthier system for everyone – brands, publishers, and consumers. That outcome is worthy of attention and passion – even if it is an uphill battle.

Welcome to The Trust Web.