TMW #136 | An internet of silos

Welcome to The Martech Weekly, where every week I review some of the most interesting ideas, research, and latest news. I look to where the industry is going and what you should be paying attention to.

👋 Get TMW every Sunday

TMW is the fastest and easiest way to stay ahead of the Martech industry.  Sign up to get the full version delivered every Sunday for this and every TMW, along with an invite to the TMW community. Learn more here.

An internet of silos

Do you remember when all websites looked like this?

This beautiful example of early 90’s web design (or lack thereof) is how people used to find things on the web. By manually curating lists of website pages that would take you to online shopping sites, blogs, forums and education, Yahoo was one of the first companies on the internet to attempt to aggregate the steadily growing volume of new information added to the web each and every day.

This was before personalized product recommendations, targeted advertising, newsfeeds, and social media, and when a university project called strangely called Backrub developed a search algorithm that eventually became Google.

Today, with the larger shift to an internet without third-party cookies and increasing data privacy regulations looming over us, something is happening that perhaps nobody could have predicted: That the internet, particularly from a marketing perspective, is becoming more siloed.

We’re going back to the web of the 90s, but this time it’s with trillion-dollar corporate hegemonies, billions in ad spend each year, and two-thirds of the world using it.

The algorithmic web

Why do I think that the web will become more siloed as we speed toward 2030? At the turn of the millennium, there was a lot of diversity in how online advertising, socializing, commerce, and content were done on the web. In fact, it was extremely hard to find services that were reliable, for both consumers and advertisers.

But all that changed through one of the most important business model innovations of all time: The algorithm. The algorithm created a way for internet companies to create a feedback loop that makes technology smarter while providing a cycle for near-endless automated growth.

This is the paradigm shift that has unlocked untold commercial value. Here’s a very basic diagram of how the mechanics of informational algorithms worked:

The internet before the creation of algorithmic recommendation machines was a lot like a Blockbuster Video: too many choices, making finding the right products, content, and services difficult.

But not all algorithms are the same. Across social media, entertainment, commerce, and search, all of them behave differently, but all of them share the same positive-feedback cycle that makes them more and more powerful.

This is what it looks like as a social paradigm for companies like Meta and Twitter:

This is what it looks like for a search paradigm for companies like Google:

This is what it looks like for an e-commerce paradigm for companies like Amazon:

This is what it looks like for an entertainment paradigm on apps like TikTok and Youtube:

The dotted lines in the above diagrams are where the algorithms become the most powerful, and it’s always centered on bringing people back to post or consume things. Another way to think about it is the dotted-line-loops are where all the monetization happens.

It’s a mistake to think that the internet was the singular transformative change in marketing. The more powerful innovation was algorithms: equations that make products, services, and content findable, that connect us with friends, that allow advertisers to target people online, to serve us infinite feeds of content.

This is the technology that allows internet giants to suck up all of the commercial activity, our attention and time, and contributions of the web. By shrinking the web into platforms that facilitate several extremely powerful algorithms, this has led to untold opportunities for billions of people. There’s a lot of access to information, commerce, and new connections that would be otherwise impossible without an algorithmic internet.

The beauty of algorithmic internet services as Scott Galloway sees it, is that it takes something that is close to worthless individually and creates huge amounts of value by using it in the aggregate:  

“In the 1980s and ’90s, aggregating millions of individually meaningless credit card transactions gave direct marketers visibility into consumer spending patterns, letting them target consumers with unprecedented precision. In the 2000s, tech repeated the trick with voting data. Google, Yelp, and Travelocity all do something similar. They alloy disparate and independently valueless pieces of data into something worth more.”

But like the thermal runaway effect in nuclear physics, algorithms also have their own exponential chain reaction in society and in the economy. Instead of Hiroshima or Nagasaki as a target, as with the first atomic bombs, algorithmic internet companies had another target in mind – Wall Street.

Google, Amazon, and Meta have turned algorithms into monopoly machines – things that can continue to grow exponentially in size, power, and influence, making the web a more closed, cloistered, and centralized place focused on a handful of companies.

As you can see, the thermal runaway effect has the same dynamic; a novel situation where exponential growth can continue unabated.

From centralized to siloed

The benefit of algorithmically-empowered internet companies is that they paved the way for not only standardizing the experience of the web but also optimizing it to a general medium that most people can access and understand.

Algorithms are also a mechanism that, in principle, should make these companies more and more profitable as consumers, sellers and advertisers use their services more. In fact, it’s one of many forces that are reducing worldwide poverty. Access to free information, customers, socializing, and education is one of the absolute miracles of the algorithmic web.

But it’s also brought with it a lot of harm.

Supermassive algorithmic social media is making us more depressed, more lonely, and overall sadder as a society. Google is under fire from the DOJ for creating an illegal monopoly in advertising, and has been caught misleading advertisers for more than three years about where YouTube ads were placed.

Consumers are increasingly worried about their privacy, with the latest research saying that 70% of US consumers are concerned about the data that’s collected about them. Google’s search product felt like magic when people first started using it, and increasingly people are moving off of search because it’s unreliable in finding high-quality information.

New social media formats that have been late to the game of algorithmic optimization – like TikTok – have risen to the level of national security threat because of the company’s links to China and its powerful and popular entertainment-focused newsfeed.

Because of these pressures, the web is becoming a more regulated, closed, and private place. And I believe that one of the things that none of us anticipated is the second-order effect of the web becoming more fragmented again because of this. But it’s not all together, or all at once.

The siloing of the web is happening on the dual planes of company growth and experience – the options for marketers to find channels and audiences to grow their brands are growing while consumers are trying new apps and services more than ever.

The growth silo

The beauty of third-party cookies is that it’s a decentralized technology. It just lives as small bits of code on every browser on Earth. The user can control how it’s stored or to block it completely, which, although decentralized, created fertile ground for Meta’s pixel tracking and Google’s web crawlers to take this technology and build single-point entry spaces for huge advertising scale.

Third-party cookies were a huge part of Google’s ad business; it was the literal backbone of the programmatic advertising industry that Google completely dominated. The introduction of Chrome in 2008 was the last foundational piece that turned Google into an advertising giant. Across all major ad platforms on the web today, Google still reigns supreme, even as new players like Apple, Amazon and TikTok have entered the ad business over the past three years. Google hasn’t even budged:

One of the reasons Google and Meta are so large is that they can attract advertisers no one else has ever been able to effectively. Meta’s base of more than 10 million advertisers is comprised mostly of small businesses.

By creating a platform that’s reinforced by self-improving algorithms, small and medium businesses can do targeted advertising and reach customers. This is possible because these platforms created an environment of limited choice.

Historically, big brands were the only ones to afford big teams and budgets to look at more channels for their advertising, including TVC, OOH and print, along with the web. Small businesses have always struggled with choice. A single window to access the unlimited customers solved on that.  

And now that third-party cookies are going away and privacy pressure is making social and search advertising more expensive and harder to do, first-party data and websites that are rich in it are the next sources of growth for marketers.

Ground zero in the shift off third-party cookies is the broadening array of retail media platforms and walled gardens. Everyone from Walmart to Disney to AWS is building large-scale advertising products to use on their own consumer bases. To use Eric Seufert’s words: everything is an ad network now.

Amazon, Apple, Walmart, and many other first-party, data-rich environments are now battling it out to be the platform of choice for closed ecosystems advertising. This means more choice and less universality for small and medium businesses and large brands alike.

Growth in walled gardens will eventuate to a more siloed landscape of advertising choice. If you’re a brand wanting to run ads on retail media, and if you want to get scale and tap into a wide variety of audiences, does this now mean you’ll have to deal with a dozen different first-party ad platforms?

One of the largest media holding companies in the world – IPG Mediabrands – announced its new Unified Retail Media Solution this week. The approach is to solve the fragmentation of media and advertising by building a platform and a dedicated media buying team so that customers can go to them to handle ads, reporting, and strategy all in one place. This is one attempt to deal with the complexity of handling all different kinds of media networks that are cut off from each other.

If Amazon, Walmart, Carrefour, Instacart, and Woolworths have their own retail media definitions, ad placement requirements, differences in schemas, ad buying prices and reporting, then something like what IPG Mediabrands has launched makes sense. But this too is also going back to the future.

Brands used to rely on agencies to do the same thing with TVCs and print advertising – the fragmentation of media in the days before the Internet was a value proposition for agencies and a real problem for brands that want to reach mass consumer audiences. In the 80s and 90s you’d have to deal with different requirements on the Wall Street Journal and the New York Times if you wanted to get your ads in both newspapers. Today, it’s the same situation just with Amazon and Instacart, or Meta and Google.

Making the web a more open space for companies to find customers comes at a big cost – spending money and time to find channels and the risk of having to specialize in domains that may not lead to any real value. Meta and Google solved this because they were the search engine for everyone and the social network for everyone. Mastery of a small amount of ad platforms means access to more people on the web.

The beauty of big monolithic platforms was that you could go into Google or Facebook, hire a digital marketer to learn ads and run them for your business – all without having to get an agency involved. So the question remains: aren’t we just going back to the way things always were as the internet fragments into smaller gardens with higher walls? It seems so.

The consumer silo

Another angle for this is the consumer experience. X - formerly known as Twitter – is unbundling into smaller niche communities like Threads, Post, Substack Notes, Mastodon, and many others, even including the recently announced text-based social experience on TikTok. And to many who’ve built an audience on X, this is a very frustrating experience. Having to build and maintain multiple social apps with different social rules is just not how people want to live their lives. The convenience of Twitter, like Facebook or Instagram was that all of your friends and connections were there.

But for consumers, more choice abounds. People are using TikTok or ChatGPT for search. Shopify did record GMV at the end of 2022 by providing a foundation for millions of small and large online stores, the complete opposite of what Amazon tried to do by centralizing all e-commerce onto its platform. Google search alternatives like DuckDuckGo have close to 100 million daily active users. You could even argue that with the rise of platforms like Substack, more people are opting for individual newsletters and blogs instead of relying on a handful of news publishers.

So, consumer choices are fraying and going in a bunch of different directions. And this will make it more challenging to form social connections online, or even find new ideas, products, or services. Dror Poleg highlights this dilemma well:

“These developments have momentous consequences. More of our interactions are happening online, and a growing share of our online interactions are mediated by algorithms. This means we gradually lose control over our attention and who we interact with. And to make things worse, the algorithms constantly change, making it harder for us to "master" them and curate our feeds. This is tough on consumers and equally tough on content producers who struggle to build an audience or lose touch with the audience they already have.”

But I would posit that this is just the natural ebb and flow of how the internet works over a long-time span. We’ll have seasons of consolidation, as we’ve just seen from the past 15 years of internet history: the web started as a fragmented place with unlimited choice, algorithms concentrated that choice to a few key platforms, and now consumers and companies are going back to a more fragmented pattern for commerce and content.

Hard reset

Marketers and advertisers are always stuck somewhere in the middle of the platform and the consumer. But this is a new beginning for a lot of marketers. The power of algorithms does not only keep their owners secure at the top of the food chain but also turns their early users into incumbents. Try competing in SEO for “best CRM software” and you’d be hard-pressed to get any traction against Hubspot and Salesforce in Google search rankings unless you want to spend a lot to get there – an unattainable goal for most start-ups.

The long tail of the internet is one of accumulating awareness, power, and influence for those who were lucky enough to start when, for many of these platforms, there were very few rules and limitless opportunities to grow.

The pressures for breaking down Big Tech are real. Heavy-handed regulation, lawsuits, and fines for the internet’s incumbents, the destruction of third-party cookies as the only open and distributed form of tracking, and even the slowness of enterprise internet companies that seem to kill more product ideas than they keep might be the best things to have happened to the internet in a long time. Only for the fact that in some of these areas, the online playing field is reset and marketers are forced to be creative once again, instead of paying Google or Meta to grow their businesses for them.

A new canvas awaits marketers who want to find new opportunities in new channels, new ideas, new audiences, and new kinds of data. But there will be pain. You will need to place your bets wisely, do your research, and see things that no one else is noticing.

Despite all the frustration that comes with increased choice, greater complexity and going back to the “old ways” of building one-to-one relationships with media businesses and channel partners, the alternative is an internet that rewards illegal monopolies, destroys competitors, makes our young people more depressed, and increases the costs for doing marketing online.

The siloed web can also be a fresh start in a post-third-party-cookie internet. But first, there must be a hard reset, and it won’t be easy.

Stay Curious,

Make sense of marketing technology.

Sign up now to get TMW delivered to your inbox every Sunday evening plus an invite to the slack community.

Want to share something interesting or be featured in The Martech Weekly? Drop me a line at juan@themartechweekly.com.