Last week, I wrote about how AI is collapsing the cost of certain marketing tactics to zero, and how the things that retain value are the ones that still require taste, judgment, and effort. This means we, as marketers, have to constantly look around the corner in an attempt to identify what is up next. Most days, I find this thrilling. I am grateful to be operating in a time where it isn’t enough to simply apply existing playbooks or juice budgets to see success.
Of course, AI makes this more difficult. The digital landscape, where most marketers have spent the majority of their time over the past two decades, is changing faster than ever before. And, as it turns out, AI is reasonably good at a lot of the tasks that we used to be responsible for: writing copy, identifying target buyers, generating images. Reasonably good is fine for most of marketing, the same way it's fine for ordering a beer at a busy bar when you really wanted a cocktail. The challenge for marketers today is identifying where 90% isn't good enough, and where the last 10% is something people will actually pay for, either with money or with attention.
I’ve been thinking a lot about curation as the next frontier of opportunity. On the surface, curation is a killer use case for AI: it's just an algorithm applied to a large dataset, which is one of the things AI does best. Curation has been a core feature of the internet for decades: it’s what made Tumblr good, and what makes Twitter bad now. Algorithmic curation is also what lets us spend less on audiences and more effectively reach the right audiences. Ad platforms have gotten dramatically better at this as they've used AI to parse through first and third-party data, intent signals, and behavioral patterns.
But interestingly, the more algorithmic curation becomes, the less I find myself relying on it. I can’t stand my Twitter feed, but I am subscribed to more newsletters than ever before and often refer to the writers as though they are my friends. I genuinely couldn’t fathom taking a restaurant recommendation from Claude, especially when my Google Maps is teeming with pins that I’ve painstakingly collected from friends and strangers alike. I love Pinterest, but I have made a purchase from a sponsored post. Since moving to DC, I’ve delighted in the experience of asking new friends for recommendations for primary care, dental, therapy, dermatologist, haircut, botox, etc. Zocdoc could never.
In all of these cases, it’s not a matter of the curation being more “accurate,” it's that the recommendations come with taste, point of view, and something at stake. While AI curation optimizes for engagement, or relevance, or some proxy, human curation optimizes for "I would feel weird if I sent you somewhere bad." Those are different functions, and the second one is hard to replicate because it requires the curator to be accountable to me in a way an algorithm structurally can't be.
A better recommendation engine doesn't get you closer to your friend's restaurant rec, because the value of your friend's rec isn't that it's more accurate. It's that she ate there, she knows you, and she'd feel weird sending you somewhere bad.
As AI collapses the cost of generic curation to zero, the only curation worth paying for (or paying attention to) is curation that carries someone's taste and reputation. We can see this playing out clearly with events as a marketing channel. Following the post-COVID event boom, people (especially those in SF, NYC, LA, Austin, etc.) became much more selective about what events they would attend. The annual tech weeks went from a hot ticket to a necessary evil to something to be avoided altogether, simply because both the events and attendees lack thoughtful curation. In contrast, intimate event experiences with curated guest lists are on the rise.
The thing that makes these events work is not the venue or the food or the talking points. It's that someone has decided who's in the room, and that decision is legible to everyone there. It’s the idea that each attendee is being hand-selected by the host because their attendance will make the event better for everyone else. You're at the table because the host wanted you at the table. Personally, I’d love to receive more event invites where the host calls out 2-3 attendees I’d get value from connecting with.
So who’s throwing the next party?
