Every two weeks in Zero Clicks is sent to over 10,000 marketers, agencies, creators, publishers and technology providers who are building the next era of digital commerce. If you would like to reach this audience and support great content, Zero Clicks needs a partner. Please reach out to [email protected] to discuss how we can build useful content together.

This newsletter explores the interplay of AI, media and commerce. Here’s your TL;DR for what awaits this week: 

• The big picture: The future of AI answer engines with Ben Grosse from Profound 

• Job posts: Outfitting millions of dad bods across America 

• The meme(s) are the message: How to save the global economy  

• One final read that’s worth your time: We all write like LLMs now 

Programming note: Zero Clicks will be off for a few weeks as I welcome a new little meatball into the world. We’ll be back in early June. 

If you miss Mike, Martech Record is hosting events that bring together commerce and content teams at brands and publishers, agencies & content creators for the purpose of working together to build a new era of digital commerce :

• Webinar: Leading commerce operator roundtable discussion. May 15th, 3pm EST. Register here. 

• Creator Roundtable. Join the creators and managers that are building the next commerce ecosystems. In the post arbitrage era commerce will become aesthetically pleasing due to the collaboration of brands, creators, publishers and technology providers. May 22nd, 12pm EST. Register here.

• Live Event: 4th Annual Marketing, Content & Commerce (MARCC) NYC: Join 250 marketers, creators, publishers, agencies & technologists in NYC from 2pm-4pm on August 4th. Register here.

We’re marketing to machines now

Zero Clicks exists because the internet that today’s marketing and media leaders grew up and made our bones on is rapidly becoming a vestige of a bygone era. The skills we used to reach our current professional stratospheres are becoming commoditized or irrelevant. We’re nothing if not in this together.
No company better encapsulates this shift than Profound, the command center for marketing leaders looking to understand and optimize how their brand features in AI answer engines…an industry that did not exist two years ago. 

If nothing else, Profound ships at an absolutely frenetic pace. A whole 48 hours after OpenAI announced adding shopping features to Chat GPT, Profound announced support for tracking GPT shopping overviews. Just cracked devs doing cracked dev things.

Profound’s worldview is fascinating – the company believes that we are all building for a future where our primary customer is not a human, but AI. I’ve never done a Q&A here at Zero Clicks, but I want to start incorporating perspectives that are bolder, weirder and more contrarian than my own, especially around AI. To that end, I caught up with Profound’s growth leader, Ben Grosse, bombarded him with questions on what marketing to our AI overlords will look like and asked if it’s really any different than conventional SEO.

Before we get to the interview, a massive hat tip to Zero Clicks reader and newly appointed Writer VP of Retail Ranjan Roy for the cleanest language around what the hell to call this new industry. While I’m partial to language model AI optimization (LMAO), “generative engine optimization” (GEO) is definitely the cleanest. 

Here’s my full conversation with Ben:

Mallazzo: Let’s start with a bang(er). On a Reddit AMA, Perplexity CEO Aravind Srinivas recently said he’s not sure if SEO will be a “thing” here, referring to his answer engine. Is he right to be skeptical that any purposeful optimization to the end of better indexing in AI driven search is possible?

BG: Performance across AI Engines will certainly look different than traditional SEO. So, in some sense, he is right. SEO as we know it may not exist across his answer engine or others. However, controlling your presence in AI answers and conversations is certainly something that will exist.

 The rules based order of optimization and performance is rapidly deteriorating. Digital and growth marketers are now tasked with grabbing the attention of probabilistic, dynamic, and opaque machines that are capable of more than rules-based analysis and retrieval. They are also now tasked with executing that across top, middle, and bottom of funnel queries. The new arbiters of answers and information have preferences, habits, processes, etc. that help them deliver on this new form factor of discovery, research, and human and machine interaction.

Consumers are using AI Search Engines to get better answers faster, and these new platforms they are using have the goal of delivering better answers, faster! With what we are seeing, there is and will continue to be levers that marketers can pull to ensure that their digital presence helps LLMs deliver on that goal. As presently constructed, the internet is mainly designed for human attention. On a basic level, Profound is helping redesign it for agents.

From the content creator (hi, it me) perspective, who will be the winners in the shift from conventional search traffic to LLMs, and by extension the shift from “creating” for humans to creating for agents? Who will be the major losers?

BG: Much is left to be determined in this arena, but based on the 1 million answers and ~20 million citations Profound is analyzing everyday, Youtube, Reddit, and Linkedin are three platforms that hold significant citation influence, especially across Perplexity, Google AIO, and ChatGPT. Other platforms like X or Instagram have more stringent gating and publisher hurdles, which make them less relevant from a citation standpoint. But X is specifically tied to Grok and Instagram is now tied to Meta’s AI application.

There will be opportunities to win across all major social and publishing channels. Taking a measured approach to citation authority will be the best formula for improving your relevance and visibility. That being said, if I had to pick one winner it would be Reddit, given its common citation relevance across multiple AI Search Engines. And if I had to pick a loser, it would be TikTok, but that is also partially a function of its broader predicament.

Great, more LinkedIn thinkbois…In your blog on how AI answer engines will transform the future, you introduce a fascinating concept that in the future people “may just be paid to think” in the form of generating vetted, expert content for answer engines. Naturally, the cynic in me says, “oh sick, that will value my work at a fraction of the rate of what it is today.”

Tell me why I’m wrong and should be more optimistic as a content creator. 

BG: The key aspect within this proposition is expert! With the swell of content creators we have seen over the past five years, the pendulum has swung to where everyone is an expert in name, but not always in practice. Profound sees a world in which AI systems will be increasingly adept at filtering out sources and content based on its validity, relevance, evidence, platform, and a variety of other key factors.

 As a content creator, you should be worried if you are missing the mark across relevance, accuracy, specificity, etc. These are critical aspects of content that will determine if machines decide to cite your content. What answer engines value and what they deliver to individuals will change over time, so the work that is valued and how it is valued will without a doubt shift.

Well my content is mostly memes, vibes and tortured Bruce Springsteen references. That and $2.90 will always get me on the subway!

Looking beyond the early hiccups (i.e. the glue on pizza era), will expertise be more or less valued than it is by Google’s blue links today? Said another way, will answer engines be better or worse at filtering out slop?

BG: Answer engines will undoubtedly get better at filtering out “slop.” AI Engines are already very good at parsing out relevant information from anywhere from five to 30 sources and building cogent and cohesive answers. Now, when they are citing that many sources, they aren’t immune to citing AI or human-generated content that is over-engineered to be cited in AI answers and conversations, but this likely won’t be the case for long. They already do a pretty good job at identifying the most relevant information from that content, but it is clear that they are nowhere near perfect. 

 With Gen AI making it far easier to produce low to medium quality content, there has been a trend towards over-indexing on producing AI slop. I have a feeling that as AI Engines get better at information retrieval, answer delivery, and proactive suggestioning, digital and content marketers have the opportunity to sharpen their strategies. I think the most likely outcome is that strategies will include a mix of both high volume content generation and ultra specific, high-signal content execution.

In a similar vein, to narrow the problem set, can marketers pick one AI-engine (i.e. Google AI Overviews) to focus on and assume enough best practices hold across the landscape?

Or are each of the engines enough of their own special snowflakes to necessitate a bespoke strategy? 

BG: AI Engines each have unique algorithms and often query entirely different search indexes, so there is no such thing as a “one size fits all” for optimizing across all the different AI Engines. Perplexity is drawn to UGC like Reddit, Google AIO gravitates to multi-modal platforms like Youtube, and ChatGPT is partial to authoritative sources like Wikipedia. We recommend to focus on the platforms that matter to you and that show the greatest opportunities for growth and improvement and to implement data-driven strategies that are tailored to each model.

How does advertising fit into this broader game and will it mess with the magic? For now, Sam Altman is still against ads in Chat-GPT and Perplexity’s $50 CPM ambitions are barely out of the cradle.

Won’t AI engines get worse the minute that the Softbank printer stops going brrrr and they have to start monetizing via advertising?

BG: This is another multi-billion dollar question! My sense is that both ads and commerce will play a major role in the evolution of AI Engines. We are already seeing this with ChatGPT rolling out commerce modalities and Perplexity’s early launch of ads and in platform purchasing. That being said, I wouldn’t expect the floodgates to open all of a sudden. 

 Sam Altman has noted his desire to pursue a more capital efficient business model without having to flip the ads switch. I’m hesitant to believe this is feasible in the long term, especially as competition continues to ratchet, consumer and enterprise spending falls deeper into limbo, and capex costs rise due to swirling geopolitical tensions. I would expect them to first move towards a take rate-based commerce enablement function, which he has already hinted at.

 Ads are likely inevitable, but bad ads are not inevitable. So far, AI Engines have been able to create magical customer experiences, both for their enterprise and consumer users. Introducing poorly designed ads poses a major risk to those customer journeys they so carefully craft. I believe this to be especially true for the AI Engines companies that are founder-led, as they will  vehemently resist any monetization lever, ads or other, that threaten any facet of the customer journey and value delivery that underpin the ethos of their company.

Expect more experimentation with low-volume, high-signal ads. The memory capabilities of Perplexity and Open AI’s o3 are remarkable and offer the potential of a transformative personalized ad experience. That being said, optimizing around answers will likely be a brands best bet at winning across AI Search for the coming years.

C’mon, Ben. “Take rate based commerce enablement” function is pretty tortured language to rebrand “affiliate marketing” for the Sand Hill Road crowd.
When you think about your broader vision, are you building software that helps brands optimize for humans viewing content with assistance from AI or truly end-to-end flows where an agent interacts with content without a human in the fold at all?

Which do you think is a bigger market opportunity? 

 BG: The overarching theme under which Profound operates is the idea that marketers are no longer catering towards human eyeballs, but towards machines. This difference is the most important structural shift in the world of AI Search. AI models are determining what citations are used, they are determining the copy within the answer, they are now determining which products you can readily buy, and we are moving to a world where they determine the actual searches as well. 

 Another interesting wrinkle is that machines are now beginning to ask the questions. With OpenAI’s Deep Research and Claude’s Computer Use as two clear examples, humans are now tasking agents with completing a wide variety of tasks. If you task Deep Research to help you build a travel schedule, the agent is responsible for executing 5, 10, 50 queries that will help them execute on that task. The way in which machines or agents ask questions will likely be different than how humans ask questions.

Choosing to cater to one approach simply isn’t an option if you want to see success. Building a holistic digital presence that specifically caters to machines, in all facets, should be a goal that brands must begin to strive for.

Ok, let’s make sure our readers walk away with an immediately actionable insight – we’ve already rebranded "affiliate marketing” here, are we just rebranding SEO best practice as well?

What’s one tactical thing I should be doing differently than how I was operating in say, the purely Google-centric SEO landscape of two years ago?

BG: Profound ran an experiment and found a 12% overlap between ChatGPT's citation set and the top 20 organic Google SERP results. On top of this, 95% of AI Search citations cannot be explained by domain authority or backlink profiles. Keep in mind, all of this is specific to primarily commercial queries with some informational in the mix. We know that top, middle, and bottom of funnel queries are taking place in AI Search. When it comes to specific questions about your industry, products, company, and competitors, no SEO strategy is capable of covering this new user behavior.

Have another cool cat like Ben who I should speak to? Drop me a note at [email protected] to tell me who to interview next.  

Featured Jobs

Head of Operations, True Classic
If you asked most eCommerce aficionados what the worst possible idea for a new DTC business was, a commodity men’s T-shirt and essentials brand occupying the no-man’s land of the middle price point would probably be top of the list.

Anyway, True Classic has bootstrapped to >$300M in revenue in less than five years, mastering both DTC and broader Amazon + retail distribution across a plethora of styles, colors and sizes. They’ve also built an effortlessly cool and irreverent brand that often feels like it is speaking directly to me as a guy in profoundly mediocre shape.

All hail the cult of the dad bod!  

Associate, Finance & Strategy, Gruns
If True Classic played the game on hard mode, Gruns is a case study in trying to play eCommerce on easy and absolutely mastering the final vestiges of the 2010s/early 2020s growth playbook. It’s AG1 as a cute gummy bear with sublime funnels and absolutely picture perfect performance marketing.

Gruns picked a wildly competitive but maximum gross margin & naturally high LTV/recurring purchase in daily-use supplements and just nailed the hell out of marketing execution. Much like True Classic they are at the forefront of every major trend in growth, currently far outpacing their peers in using AI tools like Icon to their full potential.

Perfect violation of (my) narrative that the arbitrage era is completely dead. The best always find a way.

Business Development Director, Agency, Buzzfeed

Buzzfeed currently has a market cap of $75M and is trading for well below a 1X multiple of its current revenue – tough vibes. That said, the brand still holds some gravitas among the terminally online millennials who are permanently culturally trapped in 2015 (hi!). For that reason alone, I’d look for it to get temporarily memed to the moon one or two more times by the more nostalgic members of the Reddit army.

For all the lumps, Buzzfeed has achieved success at a scale most media entrepreneurs only dream of – at its high-water mark, the publication was a ~$350M/yr business near profitability and a harbinger of enduring internet trends. And even in 2025, Buzzfeed is still very much a real business and publication with 500+ employees and several open roles at various levels. Something tells me they still have some role to play in the future of media.

Strategic Partner Manager, Wearables, Meta
If we’re all gonna end up mainlining Zuck through our eyeballs, I’d at least like us to look suave while doing it.

I’d also be remiss not to mention:  Profound has 15 open roles across sales, marketing and AI search strategy. If the above interview resonated and you think you’re a fit for any of their openings, I’d be happy to put you in touch with the team.

The great language flattening

Everything you hate about how AI writes is basically a funhouse mirror. And now, it’s turning on itself. First, LLMs picked up all of our worst impulses and ticks as communicators. Now we’re picking up theirs.

The end result is that we all have to read a hell of a lot more text than we should. This article from the Atlantic is an absolutely fascinating look at how we build tech and then it builds us.