This is a past send for The Outpost, a weekly newsletter filled with realtime insights and tactics for building more impactful customer marketing programs. To get The Outpost in your inbox every week, subscribe right here.
I just landed back from a cruise with the fam (5 days sailing the Caribbean with 2 kiddos under 4 and fighting the very real “ginge tinge” battle against the sun is only slightly less work than… well, work itself) and I’m catching up on all the things I missed. Top of my list was to watch Christy Roach’s (AirOps CMO) Outpost session on why customer marketers need to care about AEO, and today I’ve got some thoughts I need to share with my fellow in-the-trenches customer and product marketers.
I’ll admit that when AEO buzz first started taking over my LinkedIn feed last year, I mostly filed it under “things someone else in marketing is probably worrying about.”
I have enough acronyms in my life, TBH.
That lasted until my final few months at my last company, when I got roped into a marketing-team-wide push to get us cited more often in AI answers, with our VoC and advocacy programs as a big part of the plan. Made sense to me. Of course our good friends Claude, Gemini, and ChatGPT would rather hear what customers are saying about us than sift through our own marketing babble (can’t say I blame them). That push is why I ended up helping build an Awards & Recognition page for the insatiable LLMs. What Christy’s session did was put words to something I’d only half figured out back then: even if we don’t own the AEO strategy, customer and product marketers already supply a lot of what makes it work.
The reason is fairly simple: AI needs proof.
When someone asks ChatGPT, Claude, Perplexity, or another AI tool which vendor they should consider, the model has to figure out whether the claims those vendors make about themselves are actually credible. Christy described it as looking for verifiable proof, or evidence that exists on your site but can also be corroborated elsewhere.
Customer marketers produce an absurd amount of that evidence.
Customer quotes. Reviews. Case studies. Interviews. Survey data. Community conversations. Webinar transcripts. The random nugget a customer drops 23 minutes into a CS call that explains your product better than any messaging doc ever could.
And the closer someone gets to a buying decision, the more that evidence seems to matter.
AirOps analyzed more than 500,000 citations across customer journeys and found customer proof was 3.5x more likely to appear in decision-stage AI queries than early-stage queries.
Let that marinate for a second.
We’ve spent years talking about customer proof as something that helps people trust a company. Now the tools they use to research those companies are looking for the same thing. Might be a good week to forward this one to your CMO. Something tells me they’re about to take a lot more interest in your customer proof library.
I was also incredibly relieved to hear Christy warn that the answer isn’t “go make 100 more case studies.” I have a love-hate relationship with them, but that’s a subject for another newsletter.
Here’s a handful of tactics she recommends instead:
- Mine customers for frontier knowledge. Instead of only asking questions that fit the asset you already know you want to create, look for information that doesn’t exist elsewhere yet: novel use cases, original data, strong opinions, unexpected results, and the actual language customers use to describe their problems.
- Give AI corroboration outside your own domain. Your website can say you’re wonderful until the servers overheat. But AI, quite frankl, isn’t going to take your word for it. It’s looking at places like review sites, communities, forums, and other third-party sources to see whether anyone else agrees with your puffery.
- Refresh what you already have. AirOps found content published within the previous three months was 3x more likely to be cited. That said, this doesn’t necessarily mean starting from scratch on a three-month cycle. Updating an older customer story with a new quote, result, or perspective can give a good piece of proof its second wind.
- Don’t get too precious about format. Christy’s team initially saw video performing much better in their research, but once they removed customer hubs from the analysis, video-led and quote-led proof were much closer in performance. Her advice was to basically do both. Different formats create more ways for useful customer evidence to surface.
Christy gets way deeper into the research, tactics, and examples in the full conversation with Jillian. HIGHLY suggest you watch the replay here.
🔥 Campfire Chatter
Some of the best nuggets of wisdom happened in the chat. Here are a few standouts:
- Talal S. asked what actually counts as “refreshing” old content for AEO purposes, beyond rewriting some headlines, changing the publish date, and calling it a day. Christy recommended looking for substantive ways to make the information itself more current. This could mean restructuring older content for AI search, updating outdated data, or revisiting a customer featured in an older case study to get a new quote, new insight, or updated result. That great customer story from two years ago might just need a “where are they now?” section.
- Paige J. asked whether hosting customer videos on Vimeo versus YouTube makes a difference for LLM visibility. Christy recommended putting them on YouTube, even if that’s not where you embed them on your website. YouTube content is much more likely to surface in LLMs, especially Google AI Overviews. But Christy prefers Vimeo for the cleaner on-site viewing experience, so her team uses both. Vimeo gives humans a better experience on AirOps’ website, while YouTube gives AI systems another place to find the proof.
Tried any of Christy’s tactics already? Let me know how it’s going! Would love to feature you in a future Outpost session or newsletter. Now if you’ll excuse me, I have a mountain of post-cruise laundry to climb.
