My current delusion is that I, an extremely risk-averse person who is still consistently masking in 2026, would enjoy Olympic skeleton, a sport that involves hurtling headfirst at 90 mph on an icy half-pipe.
What can I say? The heart wants what it wants.
Speaking of hurtling headfirst at 90 mph, have you been keeping up with the latest in LLMs?
We have. Here’s a roundup of resources shaping how we approach LLM visibility for our clients.
First, thanks to some fascinating research by the folks at SparkToro, we know that you can’t “rank” in LLM results.
While Google can (could?) reliably produce the same 10 blue links for a given keyword search, LLMs are erratic: they don’t consistently generate the same number of results, the same order (ranking) of results, or the same actual results from one query to the next.
HOWEVER, if you run hundreds or thousands of queries, LLMs do tend to repeat certain answers more than others. What the legit tracking tools can tell you is how often (as a percentage of likely searches) your brand appears in a given LLM’s response.
This is partly why Seer Interactive founder and CEO Wil Reynolds maintains that AI visibility is a vanity metric, meaning it isn’t consistent enough to tie to leads or revenue.
And yet.
We know that people are querying LLMs.
We know that many B2B brands are seeing at least some LLM leads and conversions.
We know that many CEOs are facing pressure from boards to increase LLM visibility.
So which is it? Vanity or bonfire?
Maybe a little bit of both. In fact, I think we’re in a deliciously complex moment, which makes it an exciting one for thinking marketers. Let’s talk about why.
So: CEOs are currently willing to pour money into LLM optimization.
This means it’s an excellent time to try new things – including things that you’ve struggled to get buy-in for in the past.
And yes, the immediate goal may be increased AI visibility, but even if that doesn’t happen, you’ll likely unlock new value for your brand.
Why? Because you will try new things. You will force yourself to try unfamiliar tactics and, as a result, you will learn.
This is the gist of the second rule of improv comedy: There are no mistakes, only gifts.
Meaning, even if you “mess something up,” you’ve given yourself the gift of data about how something does (or doesn’t) work.
(I am such a big fan of this rule that I wrote an entire novel about it, for which I racked up 180 rejections. What did I learn? It’s more viable as a mindset reframe for marketers.)
If I were being pressured by my CEO to boost AI mentions right now, here’s what I would do:
For example, YouTube mentions are the metric most correlated with LLM visibility right now and Reddit is a top-cited domain. If you’ve got video content that isn’t on YouTube, why not add it over there? Why not get active in relevant Reddit communities?
If you see modest success, you can dig in to optimize (here’s a great primer on what to expect on the YouTube front). If you don’t, that’s still useful data. Now you can confidently and intentionally exclude YouTube or Reddit from your strategy.
See that? No mistakes. Only gifts.
Even better: this kind of strategy is work that can’t be outsourced to AI, which is why Business Insider is hailing…
First, WSJ declared “storytellers” the hot new thing. Now, Business Insider has noted that many big tech firms are hiring comms professionals – and paying a pretty penny, too.
The reason all these “AI will take your jobs” companies aren’t letting AI do all their comms work is that they’re not looking for high-volume commoditized writing.
They’re looking for nuance and strategy. They’re looking for judgment and taste.
They’re looking for people with the creative muscle needed to dream up experiments, the skill to execute, and the pragmatism to assess results and adjust as needed.
They’re looking, in other words, for people who can hurtle forward at top speed while staying on track and on message – and who make it look like fun. (Which is making me think maybe my skeleton dreams aren’t entirely inexplicable?)